Detailed Description
The foregoing description is only an overview of the present application, and is intended to be implemented in accordance with the teachings of the present application in order that the technical means thereof may be more clearly understood, and in order that the above-mentioned and other objects, features and advantages of the present application may be more readily understood, the following detailed description of the present application.
The present application will be described in further detail below with reference to the accompanying drawings, in order to make the objects, technical solutions and advantages of the present application more apparent, and the described embodiments should not be construed as limiting the present application, but all other embodiments obtained by those skilled in the art without making any inventive effort are within the scope of the present application.
In the following description, reference is made to "some embodiments" which describe a subset of all possible embodiments, but it is to be understood that "some embodiments" may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict, the term "first\second" being referred to merely as distinguishing between similar objects and not representing a particular ordering for the objects. The terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, apparatus, article, or server that comprises a list of steps or elements is not necessarily limited to those steps or elements that are expressly listed or inherent to such process, method, article, or apparatus, but may include other steps or modules that may not be expressly listed or inherent to such process, method, article, or apparatus, and unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. The terminology used herein is for the purpose of describing embodiments of the application only.
The embodiment of the application provides a weighing monitoring and early warning method for an intelligent scale, as shown in fig. 1, the method comprises the following steps:
and S100, placing the target intelligent scale on a detection platform, activating a smart scale detection program to carry out weighing record calculation on the target intelligent scale, determining an initial weighing coefficient, and carrying out dual backup on the initial weighing coefficient, wherein the dual backup comprises local storage and cloud backup.
Preferably, the smart balance to be detected and calibrated is placed on the detection platform, the detection platform usually has certain stability and accuracy requirements, so as to ensure that the relationship between the output signal of the smart balance sensor and the actual weight is not interfered by external factors in the weighing process, ensure the reliability of the weighing result, activate the smart balance detection program to perform weighing record calculation on the target smart balance, specifically, the smart balance detection program controls the smart balance to perform multiple operations, including checking self hardware, initializing a sensor, weighing record calculation and the like, wherein the smart balance can perform weighing operation on a standard weight object placed on the smart balance, and data obtained by weighing each time are determined through analysis and calculation of the weighing data, the initial weighing coefficient of the smart balance reflects the relationship between the output signal of the smart balance sensor and the actual weight and is used as a reference for judging whether the smart balance is accurately weighed or not, for example, a relational expression is obtained by performing weighing by using the electrical signal values output by the standard weight and the sensor, wherein the relevant parameters are initial weighing coefficients, then performing dual backup on the initial weighing coefficients, including local storage and backup, wherein the local storage is to obtain the initial weighing coefficients by the calculation, the initial weighing coefficients are stored in a cloud storage device or a cloud storage device, and the cloud storage is conveniently stored in a shared network, and the cloud storage is realized at any time, and the initial storage is convenient, and the initial storage is realized, and the cloud storage is realized, and the initial storage is easily and the fault-shared by the cloud storage is stored by the cloud storage or the cloud storage is stored, even if the local storage equipment of the intelligent scale is damaged, the loss of the initial weighing coefficient can not be caused, and the safety and the reliability of data are improved.
Further, the step S100 further comprises a step S110 of starting a smart scale detection module set based on the smart scale detection program, wherein the smart scale detection module set comprises a calibrator placing module, a weight recording module and a coefficient calculation module, a step S120 of numbering a standard weight set to obtain calibration number information, the standard weight set is sequentially placed on the target smart scale according to the calibration number information through the calibrator placing module, a step S130 of sequentially recording an actual weight set and a display weight AD value set of the standard weight set based on the weight recording module, and a step S140 of carrying out regression fitting calculation on the actual weight set and the display weight AD value set through the coefficient calculation module to determine the initial weighing coefficient.
Preferably, the intelligent balance detection module set is further activated according to the intelligent balance detection program, wherein the intelligent balance detection module set comprises a calibrator placement module, a weight recording module and a coefficient calculation module, each module is responsible for different detection tasks, detection of the intelligent balance and determination of an initial weighing coefficient are cooperatively completed, specifically, the standard weight set is a group of objects with known accurate weights and is used for calibrating the intelligent balance, the standard weights are numbered to obtain calibration number information so as to determine the identity and placement sequence of each standard weight, then the calibrator placement module sequentially places the standard weights on the target intelligent balance according to the calibration number information, the placement sequence is fixed and traceable, the weight recording module performs corresponding recording operation after the standard weights are placed on the intelligent balance, the actual weight of each object in the standard weight set is recorded, meanwhile, the weight recording module also records the weight AD value corresponding to each standard weight, which is displayed by the intelligent balance, the AD value is obtained by converting analog signals output by the intelligent sensor into digital signals, and the current weight value of the intelligent balance is displayed by the intelligent balance, and the AD value is displayed by the digital balance.
Preferably, the final coefficient calculation module performs regression fit calculation on the actual weight set and the display weight AD value set obtained by the previous record, that is, by analyzing the relationship between the two sets of data, finds an optimal mathematical model to describe the correspondence between them, for example, it is possible to use a linear regression method to find a linear equation, such as y=kx+b, where y is the actual weight, x is the display weight AD value, k is the slope to be determined, and b is a constant, to fit the two sets of data, and determine, by calculation, the initial weighing coefficient that can achieve the optimal match between the display weight AD value and the actual weight of the smart balance. For example, table 1 provides a set of data showing the different display weight AD values and corresponding actual weights, as well as the initial weighing factors (slope k and intercept b) calculated by linear regression, and verifies the accuracy of the linear equation.
Table 1 example table of linear regression fit for weighing data of smart scale
The equation y=0.97x+1 obtained by linear regression can be better fit to the relation between the display weight AD value and the actual weight, where k=0.97 and b=1 are the determined initial weighing coefficients, and then the initial weighing coefficients are used to determine the weighing accuracy of the intelligent scale, for example, a new display weight AD value is 600, and the new display weight AD value is substituted into the equation to obtain the theoretical actual weight y=0.97x600+1=583, which indicates that the intelligent scale may have the problem of inaccurate weighing if the actual measured weight deviates greatly from 583.
Further, step S100 further includes, when the initial weighing coefficient is abnormal, issuing the initial weighing coefficient by using a background authorization mechanism.
Preferably, if the initial weighing coefficient is abnormal, the initial weighing coefficient is issued based on a background authorization mechanism, for example, the weighing equipment is calibrated by using a calibration tool such as a weight of standard weight, the correct value of the initial weighing coefficient is determined according to the calibration result, for example, when the electronic scale is calibrated by using a standard weight with high precision, the proper initial weighing coefficient is calculated and set according to the actual weight of the weight and the display value of the electronic scale so as to ensure the weighing accuracy, and the initial weighing coefficient which is issued at present is determined by referring to the historical weighing data in the past normal operation and the corresponding accurate initial weighing coefficient, for example, the initial weighing coefficient which is used as a new issuing value is compared with the initial weighing coefficient in the past normal operation under the same type of equipment or the same environment.
And step 200, constructing a weighing scene test parameter table, monitoring and obtaining a real-time weight AD value set for multiple times through the target intelligent scale, and recording the actual weight information of the object.
Preferably, the weighing scene test parameter table is specifically constructed to comprehensively and systematically detect weighing performance of the intelligent balance under different scenes, such as weighing of different materials in industrial production, pricing of different commodities in commercial retail, or measuring of weights of different people in medical care, specifically, test parameters (including environmental parameters, characteristic parameters of weighing objects, weighing frequency) are defined, wherein the parameters are taken into consideration in consideration that environmental factors such as temperature, humidity and air pressure can influence sensor performance of the intelligent balance, for example, the temperature range is-10 ℃ to 50 ℃, the humidity range is 20% RH to 80% RH, weighing performance of the intelligent balance under different environmental conditions is recorded, the characteristic parameters of the weighing objects include materials (such as metal, plastic, wood and the like) of the objects, shapes (regular or irregular), weight ranges (from light objects such as tablets to heavy objects such as several tons of goods), the different materials and the shapes can generate different pressure distributions on the sensor, thus influence weighing accuracy, the same objects are determined, the weighing stability is continuously changed under the condition of continuous observation of the intelligent balance under the condition of small operating time, for example, the weighing stability is continuously changed under the condition of weighing under the condition of 8 days.
Preferably, according to the programming in the test parameter table, objects with various different characteristics are weighed for multiple times under different environmental conditions, for example, in an environment with the set temperature of 25 ℃ and the humidity of 50% RH, a regular metal block weighing 1 kg is continuously weighed for 10 times, corresponding real-time weight AD values are recorded after each weighing, and along with the progress of the test, a large amount of AD value data under different scenes can be accumulated to form a real-time weight AD value set, so that the output condition of the sensor of the intelligent scale under various complex conditions is reflected. For each object used for testing, the accurate actual weight of the object needs to be determined, for objects with known weights such as standard weights, the nominal weight of the object can be directly adopted, and for other objects, high-precision reference scales can be used for measurement to obtain accurate actual weight values. And when the real-time weight AD value is obtained each time, the corresponding actual weight information of the object is recorded in detail, and the weight AD value displayed by the intelligent scale and the actual weight of the object are compared and analyzed in a corresponding recording mode, so that the weighing accuracy and performance of the intelligent scale are evaluated.
Further, the step S200 comprises the steps of obtaining an intelligent scale application target, setting a weight threshold value of a test object and test environment information according to the intelligent scale application target and specification attributes of the target intelligent scale, determining a weight interval density according to weighing precision requirements, performing multi-position selection within the weight threshold value of the object according to the weight interval density to determine an object test weight set, performing factor extraction on the test environment information to obtain a test environment association factor set, performing parameter design on the test environment association factor set based on the intelligent scale application target to obtain a test environment factor parameter set, and building the weighing scene test parameter table by determining the weight interval density according to weighing precision requirements, and performing component integration on the object test weight set and the test environment factor parameter set.
Preferably, the application targets of the smart balance are obtained, i.e. the final use scenario and purpose of the smart balance are determined, for example, the smart balance is used for weighing and pricing of goods in commercial retail, accurate weighing of materials on an industrial production line, measurement of body weight in the healthcare field, etc., the weight threshold of the test object and the test environment information are set in combination with the specification properties (such as maximum weighing range, precision grade, etc.) of the smart balance, for example, if the smart balance is used for commercial retail, the maximum weighing range is 100 kg and the precision is + -5 g, the weight threshold of the test object may be set to 0 to 100 kg, the test environment information may include the temperature range (such as 0 ℃ to 40 ℃ and the temperature change of the simulated general commercial environment), the humidity range (such as 30% RH to 70% RH), the interval density of the test weight is selected within the weight threshold range of the smart balance according to the weighing precision requirement of the smart balance, the weight interval density is smaller, i.e. more test weight points are selected if the precision requirement is higher, otherwise, the weight interval density of the test object may be set to be greater than the precision is relatively lower, for example, the precision requirement is higher, the weight interval density may be set to be greater than the threshold of 1 kg weight per commercial weight may be set for each of the test object.
Preferably, the set test environment information is analyzed in detail, the key factors are extracted, and besides the temperature and the humidity, the key factors can also comprise air pressure, vibration conditions (if the intelligent scale is used in a vibration environment), electromagnetic interference conditions (such as electromagnetic interference can be generated by large-scale electrical equipment nearby), and the like, and the extracted key factors form a test environment related factor set, for example, the test environment related factor set can be temperature, humidity, air pressure, vibration, electromagnetic interference and the like, each factor in the test environment related factor set is specifically designed according to the application target of the intelligent scale, and the parameters of each factor form a test environment factor parameter set, for example, the test environment factor parameter set can be at a temperature of 20 ℃ to 25 ℃, the humidity of 40%RH to 60%RH, the air pressure is standard atmospheric + -5%, vibration and no obvious vibration, and the electromagnetic interference does not exist in the surrounding 1 meter. And finally, integrating and correlating the determined object test weight set with the test environment factor parameter set to obtain a weighing scene test parameter table, wherein when the weighing scene test parameter table is used for testing, the weighing test can be definitely carried out on different object test weights under different test environment parameter conditions, and further the performance of the intelligent scale under different scenes can be systematically tested and evaluated.
And step S300, carrying out partition regression fitting on the basis of the actual weight information of the object and the real-time weight AD value set, generating a weighing nonlinear model, and obtaining a dynamic weighing coefficient set according to the weighing nonlinear model.
Preferably, the intelligent balance may show different characteristics in different weight intervals, or due to various factors (such as nonlinear characteristics of a sensor, influence of a mechanical structure, etc.), the relation between the weight and the AD value is not a simple linear relation, so that the whole weight range is divided into different intervals, regression fitting is performed in each interval respectively to describe the nonlinear relation more accurately, specifically, the weight range is divided into a plurality of cells according to the actual weight information of an object and the corresponding real-time weight AD value set, the weight range is divided into a plurality of cells according to a certain rule (such as equal weight intervals, according to the data distribution characteristics, etc.), for each data in each cell, a function capable of best fitting the data in the interval is searched by adopting a proper regression method (such as a least square method, etc.), and the functions such as a polynomial function, an exponential function, a logarithmic function, etc. are integrated together, so that a weighing nonlinear model for the whole weight range is formed, the real-time AD relation between different weights and actual weight values can be described more accurately, and the real-time fitting of the weight balance can be better. It is assumed that the weighing range of 0-100 kg is divided into three sections of 0-20 kg, 20-50 kg, 50-100 kg. Within the interval of 0-20 kg, the fitted function isIn the interval of 20-50 kg, the function is(Possibly approximately linear in this interval) and a function of 50-100 kgThe whole weighing nonlinear model is a piecewise function formed by the three functions in a common range. Wherein, the、、、、、、The constant coefficients reflect the relation characteristics between the actual weight and the real-time weight AD value in different weight intervals, form a dynamic weighing coefficient set, are obtained by fitting data in different weight intervals, and change along with the change of weight. The dynamic weighing coefficient set can reflect the actual weighing condition of the intelligent scale under different weights more accurately, is beneficial to improving the accuracy and reliability of weighing, and can be used for correcting and adjusting the weighing result of the intelligent scale more accurately.
Further, the step S300 further comprises a step S310 of calculating and obtaining the actual weight information of the object and the measured weight difference information of the real-time weight AD value set, a step S320 of carrying out threshold partition on each scene parameter in the weighing scene test parameter table based on the measured weight difference information to obtain a weighing scene parameter partition threshold value, a step S330 of carrying out partition identification on the actual weight information of the object and the real-time weight AD value set according to the weighing scene parameter partition threshold value to obtain a partition measured weight data set, and a step S340 of carrying out piecewise regression fitting and combination verification tuning on the partition measured weight data set in sequence to generate a weighing nonlinear model.
Preferably, for each set of corresponding actual weight information of the object and the real-time weight AD value, the real-time weight AD value is converted into a corresponding measured weight value through a certain conversion relation (because the AD value itself may not be a direct weight unit and needs to be converted according to the characteristics of the intelligent scale), the measured weight difference information is obtained by subtracting the measured weight from the actual weight of the object, the deviation condition between the measured result of the intelligent scale and the actual weight is reflected, different threshold ranges are determined according to the measured weight difference information, each scene parameter in the weighing scene test parameter table is divided into different sections, the scene parameter in the weighing scene test parameter table is used as the same area, the different area weight data with larger scene parameter difference is obtained by dividing the sections, for example, the difference value in the range of +/-0.1 gram may be divided into one section, the scene parameter in the range of 0.1 gram to 0.5 gram may be divided into another section, and the scene parameters are classified according to different error ranges so as to analyze and process the data more carefully, and the obtained boundary values of the different sections are the weighing scene parameter partition threshold values.
Preferably, classifying and identifying the actual weight information of the object and the real-time weight AD value set according to the determined weighing scene parameter partition threshold value, classifying the data belonging to the same threshold value interval into a group to form partition measurement weight data sets, ensuring that the data in each set has similar measurement error characteristics, facilitating the subsequent processing and analysis of the data in different error ranges, fitting each partition measurement weight data set by adopting a proper regression method (such as polynomial regression, nonlinear regression and the like) respectively, finding a function model capable of optimally describing the data relationship in the set, capturing the relationship between the actual weight and the measurement weight more accurately, finally combining the segmented regression fit models, verifying and optimizing the models through cross verification, mean square error evaluation and the like, adjusting the parameters or structures of the models according to the verification results, improving the accuracy and generalization capability of the models, enabling the models to be better suitable for different weighing scenes and object weight ranges, and finally generating the weighing nonlinear model.
Further, step S320 includes step S321 of calculating mean value and standard deviation of the measured weight difference information and statistically analyzing the calculation result to obtain difference data distribution information, step S322 of selecting a target clustering algorithm according to the difference data distribution information, performing clustering analysis on the measured weight difference information by using the target clustering algorithm to obtain a difference data clustering result, step S323 of determining a weight difference data cluster according to the difference data clustering result, and step S324 of performing threshold matching partition on each scene parameter in the weighing scene test parameter table based on the weight difference data cluster to obtain the weighing scene parameter partition threshold.
Preferably, the average value and standard deviation of each measured weight difference value are calculated, wherein the average value reflects the average level of the measured weight difference value, the standard deviation is measured to be the discrete degree of measured weight difference value information, the larger the standard deviation is, the more scattered the data is, the worse the stability of the measured result is, otherwise, the more stable the measured result is, the average value and the standard deviation are comprehensively analyzed, the distribution condition of the difference value data is determined by combining specific data of the measured weight difference value information, for example, whether the data is normally distributed, biased distributed and the like, and the characteristics of the concentrated trend, the discrete degree and the like of the data are judged, a proper clustering algorithm, such as a K-average clustering algorithm, a DBSCAN density clustering algorithm, a hierarchical clustering algorithm and the like, is selected according to the characteristics of the difference value data distribution information, if the difference value data is distributed uniformly, the number of clusters is approximately known, the K-average clustering algorithm is more proper, the SCAN density clustering algorithm can accurately identify different clusters if the data distribution has different density areas, and then the measured weight difference value information is processed, namely, the measured weight difference value information is divided into different categories according to the similarity among the data so that the different categories have higher similarity.
Preferably, after cluster analysis, the obtained different categories are clustering results of difference data, each category can be regarded as a weight difference data cluster, representing different groups with similar measurement error characteristics, for example, one data cluster may contain data with smaller measurement errors and relatively stable, the other data cluster may contain data with larger measurement errors and larger fluctuation, the data with different error characteristics can be analyzed and processed more specifically by determining the weight difference data clusters, and finally, for each weight difference data cluster, the data characteristics and the distribution range of the weight difference data clusters are analyzed, a proper threshold range is determined, for example, for the data cluster with smaller measurement errors, a smaller threshold range is set, for the data cluster with larger measurement errors, a larger threshold range is set, and then, according to the threshold range, the scene parameters in the symmetrical weight scene test parameter table are partitioned, so that the scene parameters in each partition correspond to the weight difference data cluster with similar measurement error characteristics, and the weighing threshold value of the scene parameters is obtained, thereby ensuring the weighing performance and accuracy.
Further, step S340 further includes step S341 of performing linear regression fitting on the partition measured weight data set in sequence to obtain a partition weighing coefficient regression model set, step S342 of performing sequence combination on the partition weighing coefficient regression model set according to the weighing scene parameter partition threshold to obtain an initial weighing coefficient model, step S343 of performing verification evaluation on the initial weighing coefficient model to obtain a model partition decision coefficient, and step S344 of performing partition threshold optimization on the initial weighing coefficient model based on the model partition decision coefficient to generate the weighing nonlinear model.
Preferably, for each partition measured weight data set, a linear regression method is used to find a linear relationship between the data, the linear regression attempts to find a straight line (hyperplane in multidimensional space) so that the straight line can be optimally fit to the data points, that is, the square sum of the distances from the data points to the straight line is minimized, and then a corresponding linear regression model is obtained for each partition measured weight data set, describing the linear relationship between the actual weight of the object and the measured weight in the partition, wherein each model has a specific coefficient thereof, thereby forming a partition weighing coefficient regression model set, and each partition weighing coefficient regression model is combined according to a certain sequence according to the determined weighing scene parameter partition threshold, for example, if the partition is divided according to the size of the weight difference, the corresponding models are combined according to the sequence from small to large or from large to small, thereby obtaining an integral model comprising a plurality of partition models, that is, namely, the initial weighing coefficient model, and describing the relationship between the actual weight of the object and the measured weight in different weight intervals, thereby more comprehensively covering the whole weighing range.
Preferably, the initial weighing coefficient model is verified and evaluated, and the model partition decision coefficient is obtained through calculation, wherein the model partition decision coefficient is the fitting degree of the model to the data, the value range is between 0 and 1, the closer to 1, the better the fitting effect of the model to the data is, and the closer to 0, the worse the fitting effect of the model is. And then, according to the result of the model partition decision coefficients, the initial weighing coefficient model is adjusted and optimized, if the decision coefficient of a certain partition is lower, the model fitting effect of the partition is poor, or the model of the partition is required to be adjusted, or further improvement is carried out, such as the addition of higher-order terms to become a nonlinear model, so that the model can better fit the data, and finally, a weighing nonlinear model with better fitting effect in the whole weighing range is generated, thereby improving the precision and reliability of the weighing system.
Step S400, comparing and calculating the dynamic weighing coefficient set with the initial weighing coefficient in sequence to determine a weighing coefficient offset slope set.
Preferably, each coefficient of the set of dynamic weighing coefficients is compared with a corresponding coefficient of the initial weighing coefficients, the difference or ratio is calculated to measure the degree of difference between the two, and the offset slope of the weighing coefficients is further calculated, the set of offset slopes of the weighing coefficients representing the rate of change of the dynamic weighing coefficients relative to the initial weighing coefficients, and if the dynamic weighing coefficients and the initial weighing coefficients are considered as functions of a certain variable (such as the weight of an object or a weighing scene parameter), the set of offset slopes is the set of rates of change between these functions. By determining the slope set of the weighing coefficient offset, the change trend and the difference degree between the dynamic weighing coefficient and the initial weighing coefficient can be known, if the value of the slope set is smaller and stable, the initial weighing model is closer to the dynamic actual condition in different scenes, the model accuracy is higher, otherwise, if the fluctuation of the slope set is larger, the initial weighing model possibly has defects and needs to be further optimized.
And S500, constructing a weighing dynamic early warning mechanism, triggering the weighing dynamic early warning mechanism to carry out classified matching based on the weighing coefficient deviation slope set, determining a weighing early warning strategy, and carrying out weighing early warning control on the target intelligent scale through the weighing early warning strategy.
The step S500 further comprises a step S510 of carrying out partition early warning analysis on the weighing coefficient deviation slope set according to the weighing scene parameter partition threshold value and setting partition deviation slope early warning levels, a step S520 of determining a partition weighing early warning strategy according to the partition deviation slope early warning levels, and a step S530 of constructing the weighing dynamic early warning mechanism based on the partition deviation slope early warning levels and the partition weighing early warning strategy.
Preferably, various data including real-time weight AD value, actual weight information, weighing scene parameters (such as ambient temperature, humidity, material quality of articles and the like), dynamic weighing coefficient sets and initial weighing coefficients calculated by the real-time weight AD value, actual weight information, weighing coefficient offset slope sets for triggering early warning are continuously collected, different slope threshold ranges are set by analyzing historical data and actual application requirements, when the calculated weighing coefficient offset slope exceeds a preset normal range, a weighing dynamic early warning mechanism is triggered, for example, if the normal weighing coefficient offset slope range is set to be between-0.05 and 0.05, when a calculated certain slope value is 0.08, the early warning mechanism is started, and the early warning conditions are classified into different categories according to the different slope ranges. Common classification modes such as light early warning, medium early warning and high early warning, wherein the slope of the light early warning is between 0.05 and 0.1 (or similar relatively small deviation range), which indicates that a weighing system may start to generate some small deviation but not seriously affect weighing accuracy, the slope of the medium early warning is between 0.1 and 0.2, which indicates that the weighing deviation is obvious, and may affect some applications with higher requirements on weighing accuracy, because the reasons may include gradual reduction of sensor performance, slight abrasion of equipment parts and the like, and the slope of the high early warning is higher than 0.2 (or larger deviation value), the weighing system has serious problems, which are extremely likely to cause great inaccuracy of weighing results, immediate measures are required, and possible reasons include sensor faults, serious external interference or damage of key parts and the like.
Preferably, the weighing early warning strategies are determined according to the analysis of the early warning classification results, namely the countermeasures under different early warning, and corresponding control operations are automatically executed according to the determined early warning strategies to perform weighing early warning control, including sending instructions to the intelligent scale to enable the intelligent scale to enter a specific maintenance mode, such as stopping data acquisition, displaying early warning information and the like, sending notifications to related personnel to ensure that the related personnel know the situation in time and take action, starting standby equipment or processes and the like, and ensuring the accuracy of weighing data. The method comprises the steps of combining a weighing coefficient offset slope set and a weighing scene parameter partition threshold, analyzing the change condition of slopes in a section aiming at each weighing scene parameter partition, for example, researching the distribution and fluctuation of the weighing coefficient offset slope in a low-temperature environment section, setting different early warning grades for each partition according to analysis results, such as slight, moderate and high early warning, determining a partition weighing early warning strategy according to the partition offset slope early warning grade, possibly adopting strategies such as closely focusing on weighing data change, recording related parameters and the like when the early warning grade of a certain partition is slight, arranging a professional to check whether a sensor of weighing equipment is normal or not, whether equipment connection is stable or not and stopping part of weighing service with extremely high precision requirements for the partition with moderate early warning, and if the early warning grade is high, immediately stopping the weighing equipment of the area, comprehensively checking fault reasons, and simultaneously evaluating the weighing data accuracy of the area to see whether the weighing data need to be re-weighed or corrected. And finally integrating the determined partition deviation slope early warning level with the corresponding partition weighing early warning strategy to form a weighing dynamic early warning mechanism, further automatically judging which partition is currently located, triggering the corresponding early warning strategy according to the early warning level corresponding to the partition, and realizing dynamic early warning and management of the intelligent scale weighing process. As shown in table 2, the relationship between the weighing factor offset slope, the early warning level and the recommended measures, etc. is given by way of example:
Table 2 Intelligent balance weighing early warning and measure comparison table (offset slope according to weighing coefficient)
Further, step S530 includes step S531 of setting an abnormal number of times early warning threshold and an abnormal weight early warning threshold according to the false alarm processing requirement, and step S532 of constructing a weighing false alarm processing mechanism based on the abnormal number of times early warning threshold and the abnormal weight early warning threshold and additionally supplementing the weighing dynamic early warning mechanism through the weighing false alarm processing mechanism.
Preferably, the upper limit of the allowable times of abnormal conditions (such as triggering early warning but actually weighing is normal) of the weighing system in a certain time range is determined according to the actual false alarm processing requirement, for example, the upper limit of the allowable times of abnormal triggering early warning of the weighing system cannot exceed 3 times within 1 hour, the '3 times' are the abnormal times early warning threshold value, and can be set according to the false alarm frequency in historical data and the acceptable false alarm degree of business, and the weight numerical standard is set based on the false alarm processing requirement, and when the abnormal weight displayed by the weighing system exceeds the standard, the abnormal condition is considered, for example, the abnormal weight early warning threshold value is set to +/-5 kilograms for a general object weighing scene, namely, the abnormal weight condition is considered when the deviation of the weighing result and the normal expected weight exceeds +/-5 kilograms.
Preferably, the weighing system continuously monitors the abnormal times and abnormal weight conditions in the weighing process, when the abnormal times reach or exceed an abnormal times early warning threshold value, or the abnormal weight exceeds an abnormal weight early warning threshold value, a corresponding processing flow is triggered, and once the corresponding processing flow is triggered, for example, the automatic early warning function of the weighing system can be suspended under the condition that the abnormal times reach the threshold value, the automatic early warning function is converted into a manual rechecking mode, subsequent weighing data are checked and judged by staff to avoid continuous false alarm to influence the service, and for the condition that the abnormal weight exceeds the threshold value, professional symmetrical resetting equipment can be arranged to calibrate and check, so that the accurate operation of equipment is ensured, and meanwhile, the weighing data affected before are rechecked.
Preferably, the weighing dynamic early warning mechanism is supplemented according to the constructed weighing false alarm processing mechanism, the accuracy and reliability of early warning are optimized, after the weighing dynamic early warning mechanism triggers early warning, the weighing false alarm processing mechanism can further judge whether the early warning belongs to false alarm or not, if false alarm exists, the processing is carried out according to set processing measures, the interference of false alarm on business is reduced, if false alarm does not exist, the processing is carried out according to the original flow of the weighing dynamic early warning mechanism, and therefore the early warning and processing system of the whole weighing system is perfected, and the accuracy of weighing monitoring early warning of the intelligent scale is further ensured.
In the above, the weighing monitoring and early warning method for the intelligent scale according to the embodiment of the invention is described in detail with reference to fig. 1. Next, a weighing monitoring and early warning device for a smart scale according to an embodiment of the present invention will be described with reference to fig. 2.
According to the weighing monitoring and early warning device for the intelligent scale, the technical problems that weighing deviation and early warning hysteresis cannot be accurately monitored in time due to the fact that weighing of the intelligent scale is easily interfered by factors such as environment and long-term use in the prior art are solved, and the technical effects of improving weighing accuracy and monitoring early warning timeliness of the intelligent scale are achieved. As shown in fig. 2, the weighing monitoring and early warning device for the intelligent scale comprises an initial weighing coefficient determining unit 10, a weighing test parameter table constructing unit 20, a partition regression fitting unit 30, a weighing coefficient deviation slope determining unit 40 and a weighing early warning strategy determining unit 50.
The intelligent early warning system comprises an initial weighing coefficient determining unit 10, a weighing test parameter table constructing unit 20, a partition regression fitting unit 30, a weighing coefficient offset slope determining unit 40, a weighing early warning strategy determining unit 50, a weighing early warning strategy control unit and a weighing early warning strategy control unit, wherein the initial weighing coefficient determining unit is used for placing a target intelligent scale on a detection platform, activating a intelligent scale detection program to perform weighing record calculation on the target intelligent scale, determining an initial weighing coefficient, and performing dual backup on the initial weighing coefficient, the dual backup comprises local storage and cloud backup, the weighing test parameter table constructing unit is used for constructing a weighing scene test parameter table, acquiring a real-time weight AD value set through multiple monitoring of the target intelligent scale, and recording actual weight information, the partition regression fitting unit 30 is used for performing partition regression fitting on the real-time weight information of the object and the real-time weight AD value set, generating a weighing nonlinear model, and obtaining a dynamic weighing coefficient set according to the weighing nonlinear model, the weighing coefficient offset slope determining unit 40 is used for sequentially comparing the dynamic weighing coefficient set with the initial weighing coefficient, determining a weighing coefficient offset slope set, and a weighing early warning strategy is used for constructing a weighing dynamic early warning strategy, triggering the weighing early warning strategy, and performing early warning strategy control on the intelligent early warning target scale.
Next, the specific configuration of the initial weighing factor determination unit 10 will be described in detail. The initial weighing coefficient determining unit 10 further includes starting a smart scale detection module set based on the smart scale detection program, wherein the smart scale detection module set includes a calibrator placement module, a weight recording module and a coefficient calculation module, numbering a standard weight set to obtain calibration number information, sequentially placing the standard weight set on the target smart scale according to the calibration number information through the calibrator placement module, sequentially recording an actual weight set and a display weight AD value set of the standard weight set based on the weight recording module, and performing regression fitting calculation on the actual weight set and the display weight AD value set through the coefficient calculation module to determine the initial weighing coefficient.
Next, the specific configuration of the weighing test parameter table construction unit 20 will be described in detail. The weighing test parameter table construction unit 20 further comprises the steps of obtaining an intelligent scale application target, setting a test object weight threshold and test environment information according to the intelligent scale application target and specification attributes of the target intelligent scale, determining weight interval density according to weighing precision requirements, performing multi-position selection within the object weight threshold according to the weight interval density to determine an object test weight set, extracting factors from the test environment information to obtain a test environment associated factor set, performing parameter design on the test environment associated factor set based on the intelligent scale application target to obtain a test environment factor parameter set, and building the weighing scene test parameter table by combining the object test weight set and the test environment factor parameter set.
Next, the specific configuration of the partition regression fitting unit 30 will be described in detail. The partition regression fitting unit 30 further includes calculating and obtaining measured weight difference information of the actual weight information of the object and the real-time weight AD value set, partitioning each scene parameter in the weighing scene test parameter table based on the measured weight difference information to obtain a weighing scene parameter partition threshold, performing partition identification on the actual weight information of the object and the real-time weight AD value set according to the weighing scene parameter partition threshold to obtain a partition measured weight data set, and sequentially performing segment regression fitting and combination verification optimization on the partition measured weight data set to generate a weighing nonlinear model.
Next, the specific configuration of the partition regression fitting unit 30 will be described in further detail. The partition regression fitting unit 30 further includes calculating a mean value and a standard deviation of the measured weight difference information, statistically analyzing the calculation result to obtain difference data distribution information, selecting a target clustering algorithm according to the difference data distribution information, performing cluster analysis on the measured weight difference information by using the target clustering algorithm to obtain a difference data clustering result, determining a weight difference data cluster according to the difference data clustering result, and performing threshold matching partition on each scene parameter in the weighing scene test parameter table based on the weight difference data cluster to obtain the weighing scene parameter partition threshold.
Next, the specific configuration of the partition regression fitting unit 30 will be described in further detail. The partition regression fitting unit 30 further includes performing linear regression fitting on the partition measured weight data set in sequence to obtain a partition weighing coefficient regression model set, performing sequence combination on the partition weighing coefficient regression model set according to the weighing scene parameter partition threshold value to obtain an initial weighing coefficient model, performing verification and evaluation on the initial weighing coefficient model to obtain a model partition decision coefficient, and performing partition threshold value optimization on the initial weighing coefficient model based on the model partition decision coefficient to generate the weighing nonlinear model.
Next, the specific configuration of the weighing pre-warning policy determination unit 50 will be described in detail. The weighing pre-warning strategy determining unit 50 further includes performing partition pre-warning analysis on the weighing coefficient offset slope set according to the weighing scene parameter partition threshold, setting partition offset slope pre-warning levels, determining a partition weighing pre-warning strategy according to the partition offset slope pre-warning levels, and constructing the weighing dynamic pre-warning mechanism based on the partition offset slope pre-warning levels and the partition weighing pre-warning strategy.
Next, the specific configuration of the weighing pre-warning policy determination unit 50 will be described in further detail. The weighing pre-warning strategy determining unit 50 further comprises the steps of setting an abnormal number pre-warning threshold value and an abnormal weight pre-warning threshold value according to the false-warning processing requirement, constructing a weighing false-warning processing mechanism based on the abnormal number pre-warning threshold value and the abnormal weight pre-warning threshold value, and additionally supplementing the weighing dynamic pre-warning mechanism through the weighing false-warning processing mechanism.
Next, the specific configuration of the initial weighing factor determination unit 10 will be described in further detail. The initial weighing factor determining unit 10 further includes issuing the initial weighing factor using a background authorization mechanism when there is an abnormality in the initial weighing factor.
The weighing monitoring and early warning device for the intelligent scale provided by the embodiment of the invention can execute the weighing monitoring and early warning method for the intelligent scale provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.
Although the present application makes various references to certain modules in the apparatus according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and/or the server, and the included units and modules are merely divided according to functional logic, but are not limited to the above-described division, so long as the corresponding functions can be implemented, and in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not intended to limit the scope of protection of the present application.
The above embodiments do not limit the scope of the present application. It will be apparent to those skilled in the art that various modifications, combinations, and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the scope of the present application.