Disclosure of Invention
The invention provides an intelligent enterprise operation consultation method and system based on multi-source data, which can automatically find the optimal inventory management strategy for enterprises to use after a plurality of inventory management strategies are randomly generated by combining with the operation state of the enterprise, thereby effectively improving the adaptability and adjustment capability of the inventory management strategy and solving the defects in the background technology.
In order to achieve the above object, the present invention provides the following technical solutions: an intelligent enterprise operation consultation method based on multi-source data, the consultation method comprises the following steps:
s1: according to actual production requirements and supply chain distribution time, randomly generating a plurality of initial inventory management strategies based on fuzzy logic;
s2: the method comprises the steps of obtaining multi-source data of each initial inventory management strategy based on a large database, preprocessing the multi-source data, determining the weight of each multi-source data by using an analytic hierarchy process, and obtaining the fitness score of each initial inventory management strategy by weighting and calculating a plurality of multi-source data;
s3: selecting a plurality of parent inventory management strategies by combining the adaptability scores of each initial inventory management strategy through a parent selection method, carrying out pairwise information recombination on the plurality of parent inventory management strategies, carrying out information local modification on the parent inventory management strategies subjected to the information recombination, and updating the parent inventory management strategies subjected to the information local modification into child inventory management strategies;
S4: drawing the child inventory management strategy into a management set, repeating the step S3 for iteration, generating an overall management coefficient for the management set, and jumping out of an iteration loop and outputting all management sets when the overall management coefficient is larger than a corresponding coefficient threshold value;
S5: integrating the child inventory management strategies in all the management sets, performing final sorting according to the fitness scores to generate a strategy list, and outputting and sending the child inventory management strategy with the first sorting in the strategy list to an enterprise administrator.
In a preferred embodiment, in step S1, according to the actual production requirements and supply chain distribution time, a number of initial inventory management policies are randomly generated based on fuzzy logic, including the steps of:
s1.1: collecting enterprise production demand data including weekly product demand and supply chain delivery time data including time of receipt of material from a supplier;
S1.2: taking the production demand and the supply chain distribution time as input variables, taking an inventory management strategy as an output variable, and carrying out fuzzification treatment;
S1.3: randomly generating a plurality of fuzzy input variables, inputting the randomly generated fuzzy input variables into a fuzzy logic rule for reasoning to obtain a corresponding number of fuzzy output variables;
S1.4: repeating step S1.3, randomly generating a plurality of initial inventory management strategies, wherein each initial inventory management strategy comprises an order time, a safety inventory quantity and an inventory layout.
In a preferred embodiment, the multi-source data for each initial inventory management policy is obtained based on a large database, the multi-source data including the inventory cost of the enterprise and inventory turnover rate.
In a preferred embodiment, the weight of each multi-source data is determined using a hierarchical analysis method comprising the steps of:
Selecting inventory cost and inventory turnover rate as two criteria for determining weight, further dividing the inventory cost into order cost, inventory holding cost and stock shortage cost factors, dividing the inventory turnover rate into sales cost and average inventory quantity factors, and comparing the importance of each factor pair to generate a factor matrix;
And for the judgment matrix of each factor pair, calculating a weight vector by using a feature vector method or a maximum feature value method, and synthesizing the weight vector of each factor pair to obtain the weight of the final inventory cost and inventory turnover rate.
In a preferred embodiment, weighting the plurality of multi-source data to obtain fitness scores for each initial inventory management policy includes the steps of:
Acquiring the inventory cost and inventory turnover rate of each initial inventory management strategy, and determining the inventory cost weight and inventory turnover rate weight based on a hierarchical analysis method;
The inventory cost and inventory turnover rate are weighted and calculated to obtain the fitness score of each initial inventory management strategy, and the calculation expression is as follows:
In which, in the process,In order to score the degree of fitness,For the purpose of inventory costs,For the inventory turnover rate,、Inventory cost weight and inventory turnover rate weight, respectively, and。
In a preferred embodiment, a number of parent inventory management policies are selected by a parent selection method in combination with fitness scores for each initial inventory management policy, comprising the steps of:
Obtaining the fitness score of each initial inventory management strategy, summing the fitness scores of all the initial inventory management strategies to obtain a score total value, and obtaining the selection probability of the initial inventory management strategy after the fitness scores of the initial inventory management strategies are compared with the score total value;
Mapping the selection probability of each initial inventory management strategy to the wheel disc, wherein the selection probability of the initial inventory management strategy is in direct proportion to the occupied area on the wheel disc;
The starting program drives the wheel disc to rotate, when the wheel disc stops, the initial inventory management strategy selected by the pointer is reserved and deleted from the wheel disc, and mapping is carried out again after the initial inventory management strategy is deleted from the wheel disc each time;
And when the number of the selected initial inventory management strategies is equal to the number threshold value, updating the selected initial inventory management strategies into parent inventory management strategies.
In a preferred embodiment, generating overall management coefficients for the management sets, and when the overall management coefficients are greater than corresponding coefficient thresholds, skipping the iterative loop and outputting all the management sets, comprising the steps of:
The child inventory management strategies are divided into a management set, the operations of selection, recombination and information local modification are continuously carried out, and the generated child inventory management strategies are divided into the management set;
After each iteration, assigning weights to each sub-generation inventory management strategy in the management set based on a priority graph method and combining the fitness score, and then carrying out weighted average calculation on the weights of each sub-generation inventory management strategy in the management set to obtain an overall management coefficient;
Comparing the obtained overall management coefficient with a preset coefficient threshold value, wherein the coefficient threshold value is used for judging whether the adaptability score of the overall management set meets the requirement;
If the overall management coefficient is smaller than or equal to the coefficient threshold, judging that the overall fitness score of the current management set does not meet the requirement, continuing the iterative loop, if the overall management coefficient is larger than the coefficient threshold, judging that the overall fitness score of the current management set meets the requirement, jumping out of the iterative loop, and outputting all the management sets.
The intelligent enterprise operation consultation system based on the multi-source data comprises an initial module, an acquisition module, a weighting module, a calculation module, a selection module, a modification module, a circulation module, a jump-out module, a sequencing module and an output module;
an initial module: according to actual production requirements and supply chain distribution time, randomly generating a plurality of initial inventory management strategies based on fuzzy logic;
And the acquisition module is used for: acquiring multi-source data of each initial inventory management strategy based on a large database, and preprocessing the multi-source data;
And a weighting module: determining a weight of each multi-source data using an analytic hierarchy process;
the calculation module: weighting and calculating a plurality of multi-source data to obtain the fitness score of each initial inventory management strategy;
And a selection module: selecting a plurality of parent inventory management strategies by combining the adaptability scores of each initial inventory management strategy through a parent selection method;
and (3) a modification module: after the parent inventory management strategies are recombined pairwise, carrying out information local modification on the parent inventory management strategies after the information recombination is completed, and updating the parent inventory management strategies after the information local modification into child inventory management strategies;
And (3) a circulation module: drawing the child inventory management strategy into a management set, and repeating the steps for iteration;
And (5) a jump-out module: generating an overall management coefficient for the management set, and when the overall management coefficient is larger than a corresponding coefficient threshold value, jumping out of the iterative loop and outputting all the management sets;
and a sequencing module: integrating the child inventory management strategies in all the management sets, and finally sorting according to the fitness scores to generate a strategy list;
And an output module: and outputting and transmitting the child inventory management strategy of which the first is ordered in the strategy list to an enterprise administrator.
In the technical scheme, the invention has the technical effects and advantages that:
According to the invention, a plurality of initial inventory management strategies are randomly generated based on fuzzy logic according to actual production requirements and supply chain distribution time, multisource data of each initial inventory management strategy are obtained based on a large database, preprocessing is carried out on the multisource data, after the weight of each multisource data is determined by using a hierarchical analysis method, a plurality of multisource data are weighted and calculated to obtain fitness scores of each initial inventory management strategy, a plurality of father inventory management strategies are selected by combining the fitness scores of each initial inventory management strategy through a father selection method, after the plurality of father inventory management strategies are subjected to pairwise information recombination, information of the father inventory management strategies subjected to information local modification is updated to form child inventory management strategies, an overall management coefficient is generated for a management set, when the overall management coefficient is larger than a corresponding coefficient threshold, the child inventory management strategies in all the management sets are integrated, final sequencing is carried out according to the fitness scores to generate a strategy list, and the first child inventory management strategy in the list is output to enterprise management personnel. The consultation method can effectively generate the total optimal management strategy according to the running condition of the enterprise, thereby effectively improving the adaptability and the adjustment capability of the inventory management strategy.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1: referring to fig. 1, the method for intelligent enterprise operation consultation based on multi-source data according to the present embodiment includes the following steps:
S1: according to actual production requirements and supply chain distribution time, a plurality of initial inventory management strategies are randomly generated based on fuzzy logic, and according to the actual production requirements and supply chain distribution time of an enterprise, the current operation conditions of the enterprise can be effectively matched;
s2: the method comprises the steps of obtaining multi-source data of each initial inventory management strategy based on a large database, preprocessing the multi-source data, determining the weight of each multi-source data by using an analytic hierarchy process, weighting and calculating a plurality of multi-source data to obtain the fitness score of each initial inventory management strategy, and comprehensively analyzing the quality of each initial inventory management strategy;
S3: selecting a plurality of parent inventory management strategies by combining the adaptability scores of each initial inventory management strategy through a parent selection method, carrying out pairwise information recombination on the plurality of parent inventory management strategies, carrying out information local modification on the parent inventory management strategies after the information recombination is completed, updating the parent inventory management strategies after the information local modification into child inventory management strategies, and increasing randomness so as to improve the robustness and adaptability of the inventory management strategies;
S4: drawing the child inventory management strategies into a management set, repeating the step S3 for iteration, generating an overall management coefficient for the management set, jumping out of an iteration loop and outputting all the management sets when the overall management coefficient is larger than a corresponding coefficient threshold value, and carrying out multiple iterations to acquire a plurality of inventory management strategies, so that the coverage range of the inventory management strategies is wider;
S5: integrating the child inventory management strategies in all the management sets, finally sorting according to the fitness scores to generate a strategy list, outputting the child inventory management strategy with the first sorting in the strategy list, and sending the child inventory management strategy to an enterprise manager, so that the enterprise is inventory-managed by the optimal inventory management strategy, and the management efficiency is improved.
Example 2: according to the embodiment, a plurality of initial inventory management strategies are randomly generated based on fuzzy logic according to actual production requirements and supply chain distribution time, and the method comprises the following steps:
Collecting data: production demand data and supply chain delivery time data for an enterprise over a period of time are collected. For example, the production demand data includes weekly product demand and the supply chain delivery time data includes time of receipt of material from the supplier.
Blurring: and using the production demand and the supply chain distribution time as input variables, and performing blurring processing. For production demand, fuzzy sets such as "high demand", "medium demand", "low demand" and the like may be classified, and for supply chain delivery time, fuzzy sets such as "fast delivery", "normal delivery", "slow delivery" and the like may be classified.
Establishing a fuzzy logic rule: based on the inventory management experience and expertise of enterprises, a fuzzy logic rule base is established, and inventory management strategies are used as output variables, including 'stock in advance', 'stock without', and the like. For example, if the production demand is "high demand" and the supply chain delivery time is "quick-lead", then the inventory management policy may be to increase the order to deal with the high demand.
Randomly generating fuzzy input variables: a number of fuzzy input variables such as production demand and supply chain delivery time are randomly generated. For example, an input variable is generated that has a production demand of "high demand" and a supply chain delivery time of "normal delivery".
Fuzzy reasoning: and inputting the randomly generated fuzzy input variable into a fuzzy logic rule for reasoning to obtain a corresponding fuzzy output variable. For example, based on fuzzy logic rules, it is inferred that the inventory management policy in this case might be to stock in advance to account for supply chain delays in normal delivery.
Generating a plurality of initial policies: repeating the above steps, randomly generating a plurality of initial inventory management policies, each initial inventory management policy including an order time, a secure inventory quantity, and an inventory layout. For example, a plurality of different production demand and supply chain delivery time combinations may be generated, each corresponding to an initial inventory management strategy;
The application judges the operation condition of the enterprise by acquiring the production demand and the supply chain distribution time of the enterprise in the past period of time, thereby randomly generating a plurality of inventory management strategies matched with the operation of the enterprise;
The order time is a plurality of order time points generated by enterprises in the past period according to the production demand and the delivery time of the supply chain, and different order time points can influence the operation of the enterprises, such as changing the storage level or changing the cost; the safety stock quantity is a plurality of safety stock quantities which are generated by enterprises in a period of time according to the production demand quantity and the supply chain distribution time, different safety stock quantities can influence the operation of the enterprises, the stock layout is a plurality of layout modes which are generated by the enterprises in a period of time according to the production demand quantity and the supply chain distribution time, and the stock layout is stored in a stock management strategy in a binary coded running mode;
For example: each shelf or storage area in the warehouse is numbered and a binary number is used to indicate whether each area has cargo. For example, if the warehouse has 10 shelves, a 10-bit binary string may be used to represent the status of each shelf, 1 for the presence of a good, and 0 for the absence of a good;
The production demand and supply chain delivery times at each time point will generate a set of corresponding order times, secure inventory, inventory arrangements.
The method comprises the steps of obtaining multi-source data of each initial inventory management strategy based on a large database, preprocessing the multi-source data, and determining the weight of each multi-source data by using a hierarchical analysis method, wherein the method comprises the following steps of:
Acquiring multi-source data of each initial inventory management strategy based on a large database, wherein the multi-source data comprise the inventory cost and inventory turnover rate of an enterprise, and preprocessing the inventory cost and the inventory turnover rate, including cleaning, denoising, standardization and the like, so as to ensure the data quality and consistency;
selecting inventory cost and inventory turnover rate as two criteria to be weighted, further splitting the inventory cost into order cost, inventory holding cost and backorder cost factors, splitting the inventory turnover rate into sales cost and average inventory quantity factors, comparing the importance between each factor pair in pairs to generate factor matrix, for example, for comparison between order cost and inventory holding cost, giving the following comparison matrix:
; the comparison value indicates that the importance of the order cost relative to the inventory holding cost is 3 times, the judging methods of other factor pairs are similar, and the factor matrix generating method belongs to the prior art and is not described one by one;
For each judgment factor matrix, calculating a factor matrix consistency index and a consistency ratio, wherein the consistency ratio is generally used for measuring the consistency degree between the judgment results, and if the consistency ratio is smaller than a certain threshold (generally 0.1), judging that the judgment between the judgment results is consistent;
For the judgment matrix of each factor pair, a feature vector method or a maximum feature value method is used to calculate a weight vector. Assuming that a maximum eigenvalue method is used, for the above judgment matrix, the maximum eigenvalue is 3.25, the eigenvector is [0.8, 0.4], and then we normalize the eigenvector to obtain a weight vector of [0.67, 0.33];
Finally, we integrate the weight vectors of each factor pair to get the weights of the final inventory cost and inventory turnover, namely: Because the inventory cost and inventory turnover rate of different initial management strategies are different, the inventory cost and inventory turnover rate weight of each initial management strategy can be the same or different, and compared with the existing weight assignment mode, the adaptability of the application is effectively improved.
Weighting and calculating a plurality of multi-source data to obtain the fitness score of each initial inventory management strategy, wherein the method comprises the following steps:
Acquiring the inventory cost and inventory turnover rate of each initial inventory management strategy, determining the inventory cost weight and the inventory turnover rate weight based on a hierarchical analysis method, and then calculating the inventory cost and the inventory turnover rate weight to acquire the fitness score of each initial inventory management strategy, wherein the calculation expression is as follows: In which, in the process,In order to score the degree of fitness,For the purpose of inventory costs,For the inventory turnover rate,、Inventory cost weight and inventory turnover rate weight, respectively, and。
Selecting a plurality of parent inventory management strategies by combining the adaptability scores of each initial inventory management strategy through a parent selection method, comprising the following steps:
The father selection method comprises a roulette selection method and a competitive bidding competition selection method;
The roulette wheel wager selection method comprises the steps of:
Obtaining fitness scores of all initial inventory management strategies, summing the fitness scores of all the initial inventory management strategies to obtain a score total value, obtaining selection probability of the initial inventory management strategies after the fitness scores of the initial inventory management strategies are compared with the score total value, mapping the selection probability of each initial inventory management strategy to a wheel disc, enabling a program to drive the wheel disc to rotate, retaining and deleting the initial inventory management strategies selected by a pointer from the wheel disc when the wheel disc is stopped, re-mapping after deleting the initial inventory management strategies from the wheel disc each time, updating the selected initial inventory management strategies into parent inventory management strategies when the number of the selected initial inventory management strategies is equal to a number threshold value, wherein the wheel disc bet selection method has certain randomness, and the purpose of increasing randomness is to increase the robustness of the system when the initial inventory management strategies with higher fitness scores are easier to be selected;
The competitive bidding competition selecting method comprises the following steps:
A certain number of initial inventory management policies are randomly selected from the initial inventory management policies as participants of the competitive game. The selection may be repeated until the set number of competitive participants is reached, and the competitive is scored according to its fitness for the selected competitive participants. The initial inventory management strategy with large fitness score wins and is reserved, and the wining individual is selected from the individuals participating in the competitive competition and is used as the parent inventory management strategy. If the competitive game parameters are set to select two initial inventory management strategies, repeating the competitive game twice, selecting one winning initial inventory management strategy as a parent inventory management strategy each time, updating the selected initial inventory management strategy as the parent inventory management strategy when the number of the selected initial inventory management strategies is equal to the number threshold value, and preferentially considering the initial inventory management strategy with larger fitness score by the competitive game selection method, wherein the initial inventory management strategy is optimized in advance, but the robustness of the system is reduced;
Thus, in summary, the present application selects the roulette method to select the initial inventory management strategy.
After the parent inventory management strategies are recombined in pairs, the parent inventory management strategies after the information recombination are subjected to information local modification, and the parent inventory management strategies after the information local modification are updated into child inventory management strategies, and the method comprises the following steps:
And (4) recombining two-by-two information: two individuals are randomly selected from the parent inventory management strategy to carry out pairwise information recombination. And performing cross operation on the information such as ordering time, safety stock quantity, stock layout and the like of the two selected individuals to generate new child individuals.
Information local modification: for each generated child individual, a local modification operation is performed. The local modification may be to randomly select certain parameters for adjustment or to mutate some parameters according to a certain mutation strategy. It is contemplated that parameters such as order time, secure inventory, inventory layout, etc. may be randomly increased or decreased or adjusted.
Updating to a child inventory management policy: and adding the child individuals with the information reorganization and the local modification to a child inventory management policy set.
For example:
let us assume that we have two parent inventory management policies, which are respectively as follows:
Parent inventory management policy a: order time: every monday; secure stock quantity: 100; inventory layout: a section A shelf 1;
parent inventory management policy B: order time: three weeks; secure stock quantity: 150; inventory layout: zone B shelves 2;
now we perform pairwise information reorganization: randomly selecting two individuals from the parent inventory management strategy to carry out pairwise information recombination, and supposing that we select the parent strategies A and B;
and (4) carrying out pairwise information recombination: suppose we select the crossover point at the safe stock parameters, then the offspring individuals after crossover are:
child inventory management policy C: order time: every monday; secure stock quantity: 150; inventory layout: a section A shelf 1;
Child inventory management policy D: order time: three weeks; secure stock quantity: 100; inventory layout: b region shelf 2
The child individuals are now subjected to local modification of information: assuming that we locally modify the child inventory management policy C, change the order time to every Tuesday, increase the safe inventory to 200, and keep the inventory layout unchanged;
child inventory management policy E: order time: every two weeks; secure stock quantity: 200; inventory layout: a section A shelf 1;
Updating to a child inventory management policy: the child inventory management policies D and E are added to the child inventory management policy set.
Drawing the child inventory management strategy into a management set, repeating iteration, generating an overall management coefficient for the management set, and jumping out of an iteration loop and outputting all management sets when the overall management coefficient is larger than a corresponding coefficient threshold value, wherein the method comprises the following steps:
The child inventory management strategies are divided into management sets, the operations of selection, recombination and information local modification are continuously carried out, the generated child inventory management strategies are divided into the management sets, after each iteration, weights are given to each child inventory management strategy in the management sets based on a priority diagram method in combination with fitness scores, then the weights of each child inventory management strategy in the management sets are weighted and averaged to obtain an overall management coefficient, the obtained overall management coefficient is compared with a preset coefficient threshold, the coefficient threshold is used for judging whether the fitness scores of the whole current management sets meet the requirements or not, if the overall management coefficient is smaller than or equal to the coefficient threshold, it is judged that the fitness scores of the whole current management sets do not meet the requirements, iteration loops are continued, if the overall management coefficient is larger than the coefficient threshold, it is judged that the fitness scores of the whole current management sets meet the requirements, iteration loops are jumped out, and all the management sets are output;
For example:
After one iteration is completed, the number of the child inventory management strategies is 5 in the generated management set, and weights are respectively assigned on the basis of an order diagram method according to the fitness scores of the 5 child inventory management strategies, and the weights are specifically shown in the table 1:
TABLE 1
In table 1, five sub-generation inventory management strategies are respectively corresponding to K1, K2, K3, K4 and K5 after being ranked from large to small according to fitness scores.
Integrating the child inventory management strategies in all the management sets, performing final sorting according to the fitness scores to generate a strategy list, and outputting and sending the child inventory management strategy with the first sorting in the strategy list to an enterprise administrator.
Example 3: the embodiment provides an intelligent enterprise operation consultation system based on multi-source data, which comprises an initial module, an acquisition module, a weighting module, a calculation module, a selection module, a modification module, a circulation module, a jump-out module, a sequencing module and an output module;
An initial module: according to actual production requirements and supply chain distribution time, generating a plurality of initial inventory management strategies based on fuzzy logic at random, and sending the initial inventory management strategies to an acquisition module;
And the acquisition module is used for: acquiring multi-source data of each initial inventory management strategy based on a large database, preprocessing the multi-source data, and transmitting the preprocessed multi-source data to a weighting module;
and a weighting module: determining the weight of each multi-source data by using an analytic hierarchy process, and sending the weight information of each multi-source data to a calculation module;
The calculation module: weighting and calculating a plurality of multi-source data to obtain the fitness score of each initial inventory management strategy, and transmitting the fitness scores to a selection module;
And a selection module: selecting a plurality of parent inventory management strategies by combining the fitness scores of each initial inventory management strategy through a parent selection method, and sending the parent inventory management strategies to a modification module;
And (3) a modification module: after the parent inventory management strategies are recombined pairwise, carrying out information local modification on the parent inventory management strategies after the information recombination is completed, updating the parent inventory management strategies after the information local modification into child inventory management strategies, and sending the child inventory management strategies to a circulation module;
And (3) a circulation module: drawing the child inventory management strategy into a management set, repeating the steps to iterate, and sending iteration information to a jumping-out module;
And (5) a jump-out module: generating an overall management coefficient for the management set, and when the overall management coefficient is larger than a corresponding coefficient threshold value, jumping out of the iterative loop and outputting all management sets, and sending management set information to the sequencing module;
And a sequencing module: integrating the child inventory management strategies in all the management sets, and finally sorting according to the fitness scores to generate a strategy list, and sending the strategy list to an output module;
And an output module: and outputting and transmitting the child inventory management strategy of which the first is ordered in the strategy list to an enterprise administrator.
The above formulas are all formulas with dimensions removed and numerical values calculated, the formulas are formulas with a large amount of data collected for software simulation to obtain the latest real situation, and preset parameters in the formulas are set by those skilled in the art according to the actual situation.
In the description of the present specification, the descriptions of the terms "one embodiment," "example," "specific example," and the like, mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The preferred embodiments of the invention disclosed above are intended only to assist in the explanation of the invention. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise form disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best understand and utilize the invention. The invention is limited only by the claims and the full scope and equivalents thereof.