Driving active safety robot control system and method based on artificial intelligenceTechnical Field
The invention relates to the technical field of driving safety control systems, in particular to a driving active safety robot control system and method based on artificial intelligence.
Background
With the rapid development of the transportation industry, the number of vehicles is continuously increased, and ensuring the driving safety becomes an important task. Traditional driving safety guarantee measures mainly depend on the careful driving of drivers and the constraint of traffic regulations, but traffic accidents still occur due to uncertainty of human factors and complexity of road environments.
In recent years, the rapid development of artificial intelligence technology brings new solutions to traffic safety. An active driving safety robot control system based on artificial intelligence is generated, and aims to improve the driving safety of vehicles and reduce the accident rate through advanced technical means.
In current traffic environments, drivers may face various disturbances and risk factors during driving. For example, actions such as making a call, distraction, approaching unaware of a rear vehicle, smoking, shielding of a probe, and unbelting may all result in serious safety hazards. In addition, different road conditions, weather conditions and vehicle performance can also have an impact on driving safety. The traditional safety monitoring system often has the problems of untimely response, high false alarm rate, single function and the like, and is difficult to meet the high requirements of modern traffic on safety guarantee.
To address these issues, artificial intelligence technology is introduced into the field of traffic safety. Through image recognition, sensor technology and machine learning algorithm, the running state of the vehicle and the behavior of the driver can be monitored in real time, potential safety risks can be found in time, and corresponding alarming and processing measures are adopted. For example, various illegal actions of a driver, such as call receiving, smoking and the like, can be accurately identified by using a camera and a deep learning algorithm, and an alarm is timely sent to remind the driver to correct the illegal actions. Meanwhile, parameters such as the speed, the acceleration, the steering angle and the like of the vehicle can be monitored in real time through analysis of the vehicle sensor data, and whether the vehicle is in a safe driving state or not is judged.
In addition, with the development of the internet of things technology, information interaction between vehicles and a traffic management system becomes more convenient. The driving active safety robot control system based on artificial intelligence can utilize the information to realize cooperative monitoring and management of the vehicle. For example, when one vehicle detects that a dangerous situation occurs on a road ahead, information can be timely transmitted to surrounding vehicles through a network, and other vehicles are reminded of avoiding. Meanwhile, the traffic management department can also know the running state of the vehicle in real time through the system and timely take traffic control measures, so that the road traffic efficiency and the safety are improved.
In summary, the driving active safety robot control system based on artificial intelligence is an innovation product meeting the development requirement of the age. The intelligent vehicle safety control system fuses advanced technologies such as artificial intelligence, a sensor technology and an Internet of things technology, and provides a more reliable and efficient solution for guaranteeing vehicle running safety. In the future traffic field, the system is expected to play a more important role, and is used for travel safety and navigation protection of people.
Disclosure of Invention
The invention provides a driving active safety robot control system and method based on artificial intelligence, which are used for solving the problems in the prior art.
In order to achieve the purpose, the invention adopts the following technical scheme that the driving active safety robot control system based on artificial intelligence comprises:
the alarm level module is used for setting different types of alarm levels and setting corresponding responses, the alarm levels define the response levels of different alarms, and the different alarm levels can be started in combination with actual practice;
The alarm type module is used for setting different driving alarm types, configuring corresponding rules, configuring corresponding alarm levels and alarm attributes (call receiving alarm, distraction alarm, rear approaching alarm, smoke alarm, probe shielding alarm and unbelted safety belt alarm) according to the different alarm types, and setting corresponding alarm levels; the configuration of different alarm types is marked as different rule codes, namely S= { S1, S2, S3..SN }, then the configuration is classified according to S= { S1, S2, S3..SN }, the rule check carries out integrity judgment on the classified configuration, and the rule is ensured to be available;
The processing type module is used for monitoring the generated alarm data, setting different driving safety behavior processing modes, configuring corresponding rules, including alarm data acquisition, data transmission and introduction of an AI robot in processing, receiving alarm information data by the robot during processing, acquiring alarm level configuration through a data acquisition center, and carrying out automatic flow assembly by combining corresponding flow treatment, and sending the alarm data to a data processing center after completion so as to facilitate alarm data processing and execution;
The alarm flow module is used for setting how to treat after the alarm is generated, configuring corresponding execution rules, carrying out dynamic flow execution after receiving the data of the processing type module, carrying out real-time analysis and processing on the acquired data by the flow mainly by using a natural language processing technology and a statistical analysis method, and constructing an intention recognition model by a model construction unit based on a machine learning algorithm and a finite element analysis method, so that the alarm processing flow is convenient to accurately judge, the corresponding treatment can be responded efficiently, the algorithm can also be iterated continuously by settling the data in the driving active safety field, and the complex alarm flow is convenient to recognize and process;
The AI monitoring robot module is a control center for monitoring driving active safety behaviors, is used for setting a behavior monitoring mode, and is used for monitoring after correction, the safety behaviors in the driving can be identified according to different images by dynamically inputting an early warning rule, and are fed back to different users for prompt and alarm, when the occurrence of the safety behaviors is detected or a preset alarm rule is triggered, the early warning information is notified by an automatic short message, and the alarm rule is set and adjusted according to the actual running condition;
the AI knowledge base module is used for setting different active safety knowledge bases, classifying, pushing and early warning, can be continuously perfected based on the existing safety behavior knowledge base, realizes continuous enrichment and perfection of the AI knowledge base through the processes of receiving, processing and feeding back, and is convenient for better forming active safety monitoring;
The group information module manages user group data, identifies user identities and group members in the group, utilizes JSON format and user-defined multi-group information, formulates a time node sequence set according to the user personalized pushing requirement to identify a pushing scheme of a target data service, obtains a user personalized pushing scheme set, and carries out short message pushing management of the target data service based on the user personalized pushing scheme.
Further, the alarm level module comprises the following operation steps:
Respectively setting a set for the alarm levels, wherein the set is used for carrying out grading response according to the alarm levels and can be formed for the alarm levels;
and setting equipment alarm and platform alarm modes for different alarm levels respectively, realizing equipment alarm mode definition and platform alarm mode definition sets under different alarm levels, combining the equipment alarm mode definition and the platform alarm mode definition sets, and importing the obtained M concentrated information entropy sets into the module as alarm level information sets to finally obtain alarm level data information.
And (3) introducing an alarm level response stability formula: WhereinIndicating the stability of the response of the alarm stage,The total number of alarms is indicated and,Indicating the response accuracy (the number of response accuracy divided by the total number of times) of the ith alarm,Indicating the accuracy of the average response and,Indicating the maximum response accuracy rate and,Representing the minimum response accuracy. The formula evaluates the response stability of the alarm level module by measuring the fluctuation condition of the response accuracy rate and the relation with the maximum response accuracy rate and the minimum response accuracy rate, and ensures that the system can stably respond according to the alarm level under different conditions.
Further, the alarm type module comprises the following operation steps:
The method is used for setting different driving alarm types, configuring corresponding rules, setting corresponding alarm levels and alarm attributes (call receiving alarm, distraction alarm, rear approaching alarm, smoke evacuation alarm, probe shielding alarm and unbelted safety belt alarm) according to different alarm types, and setting corresponding alarm levels;
The configuration for the different alarm types is denoted as different rule codes, denoted S1, S2, S3. And classifying according to S= { S1, S2 and S3..SN } and carrying out integrity judgment on the classified configuration by rule verification to ensure that the rule is available.
Introducing an alarm type configuration complexity formula: WhereinRepresenting the complexity of the alarm type configuration,The number of alarm types is indicated,A weight coefficient representing the ith alarm type (determined based on the importance of the alarm type and frequency of occurrence factors),The complexity of rule configuration (which can be measured by the number of rules and the number of parameters) representing the ith alarm type. The formula comprehensively considers the weights of different alarm types and the rule configuration complexity, and evaluates the configuration complexity of the alarm type module so as to better optimize and manage.
Further, the processing type module specifically includes:
Setting different driving safety behavior processing modes, configuring corresponding rules, including alarm data acquisition, data transmission and introduction of an AI robot in processing, receiving alarm information data by the robot during processing, simultaneously acquiring alarm level configuration through a data acquisition center, and carrying out automatic flow assembly by combining corresponding flow treatment, and sending the alarm data to a data processing center after completion so as to facilitate alarm data processing and execution.
Introducing a treatment efficiency formula: WhereinThe processing efficiency is indicated by the fact that,Indicating the number of alarms that are correctly handled,The total number of alarms is indicated and,Representing the average processing time. The formula comprehensively measures the working efficiency of the processing type module by considering the proportion of the correct processing times to the total alarm times and the reciprocal of the average processing time, so that the system can be ensured to process alarm data rapidly and effectively.
Further, the alarm flow module comprises the following steps:
s1.1, receiving alarm push data, namely, transmitting a task to a positioning equipment module through JT808 unified protocol, wherein the positioning equipment module is responsible for receiving and analyzing a task instruction, and ensuring the accuracy and the integrity of task information;
s1.2, pushing related early warning and voice information to related responsible persons through a robot after the receiving equipment pushes an alarm, and pushing different templates according to different types of people;
S1.3, AI alarms, namely after the pushing task is completed, the system emergently sends an emergency alarm voice instruction to alarm equipment through JT808, detects whether pushing alarm is completed or not, and dynamically identifies relevant voices according to alarm information;
S1.4, high-frequency alarming, namely sending an instruction to alarm equipment through a robot, grabbing an active safety behavior correction state, carrying out identification verification on a server, and if the correction is not carried out, starting a high-frequency alarming flow and continuously sending a voice instruction;
S1.5, loop alarm, namely detecting the behavior after high-frequency alarm, if the behavior is not changed, starting the loop alarm, and continuously executing S1.1> S1.2> S1.3> S1.4> S1.5 until the behavior is corrected.
Introducing an alarm process integrity formula: WhereinThe integrity of the alarm flow is indicated,Indicating the number of alarm process steps that were successfully performed,The number of steps in the total alarm flow is indicated,The ratio of the average execution time per step to the maximum allowed execution time is expressed. The formula evaluates the integrity and the execution effect of the alarm flow module by considering the proportion of the number of successfully executed steps to the total number of steps and the rationality of the execution time of each step, so as to ensure that the alarm flow can be executed efficiently and completely.
Further, the AI monitoring robot module specifically includes:
the data storage unit is used for storing the data uploaded by the active safety behavior alarm data transmission unit and ensuring that the data cannot be lost;
The data analysis unit processes the transmitted data, performs real-time analysis and processing on the acquired data by using a natural language processing technology and a statistical analysis method, the model construction unit constructs an intention recognition model based on a machine learning algorithm and a finite element analysis method, learns and trains data information related to active safety behaviors by adopting a Support Vector Machine (SVM) and an Artificial Neural Network (ANN) machine learning algorithm, establishes a mapping relation between the intention and an entity, optimizes model parameters by combining the finite element analysis method, improves the accuracy and the reliability of the model, and ensures that the prediction accuracy of the model is not lower than 90% on a test set.
In the AI monitoring robot model construction step, vector conversion is carried out on the corresponding grabbing images, and behavior recognition in the images is carried out through a convolutional neural network.
Convolutional neural networks are typically stacked of convolutional layers, pooling layers, and fully-connected layers. The convolution layer utilizes a plurality of different convolution kernels to extract the characteristics of the target and generate a characteristic map, the pooling layer is used for downsampling, combining the characteristics of adjacent characteristic maps and reducing the dimension, and the full-connection layer plays a role in mapping the learned distributed characteristics to a sample marking space.
Introducing a behavior recognition accuracy rate improvement rate formula: WhereinRepresents the improvement rate of the behavior recognition accuracy,The accuracy of behavior recognition after optimization is represented,Representing the accuracy of behavior recognition before optimization. The formula is used for measuring the improvement degree of the behavior recognition accuracy of the AI monitoring robot module before and after optimization, and ensuring that the recognition capability of the system on the driving safety behavior can be continuously improved.
And formulating a time node sequence set according to the personalized pushing requirement of the user to identify a pushing scheme of the target data service, obtaining a personalized pushing scheme set of the user, and carrying out short message pushing management of the target data service based on the personalized pushing scheme of the user.
And introducing a group information push accuracy formula: WhereinRepresents the group information push accuracy rate,Indicating the number of information that was correctly pushed to the target user,Indicating the total number of pushes of information. The formula is used for evaluating the accuracy of the group information module for pushing the user information, and ensuring that the system can accurately push the related information to the target user.
Further, the driving active safety robot control method based on artificial intelligence comprises the following steps:
setting alarm levels, namely setting different types of alarm levels to form an alarm level set, defining equipment alarm and platform alarm modes for the different alarm levels, and importing alarm level information sets into a system;
Setting different driving alarm types, configuring alarm levels and alarm attribute rules according to the alarm types, classifying the different alarm types and checking the rules;
setting different driving safety behavior processing modes, configuring corresponding rules including alarm data acquisition and transmission and AI robot flow assembly, and sending the processed data to a data processing center;
The alarm process execution step is that the JT808 unified protocol is used for receiving alarm push data and carrying out emergency alarm push, AI alarm, high-frequency alarm and cyclic alarm until the safety behavior is corrected;
The AI monitoring step comprises the steps of monitoring driving safety behaviors by using an AI monitoring robot module, including alarm data storage and analysis, performing behavior recognition by using a convolutional neural network, constructing an intention recognition model, and improving the accuracy of behavior recognition;
Setting different active safety knowledge bases, and continuously enriching and perfecting the AI knowledge bases through receiving, processing and feedback processes to realize classified pushing early warning;
and a group information management step of managing user group data, identifying group user identities and members, and performing short message pushing management according to user personalized pushing requirements by utilizing JSON format and user-defined multi-group information.
The method is introduced into a comprehensive effect evaluation formula:
WhereinThe method is represented by a comprehensive effect evaluation value,Indicating the stability of the response of the alarm stage,Representing the complexity of the alarm type configuration,The processing efficiency is indicated by the fact that,The integrity of the alarm flow is indicated,Represents the improvement rate of the behavior recognition accuracy,Represents the group information push accuracy rate,、、、、、The weight coefficient is set according to the importance degree of different indexes. The formula comprehensively considers key indexes of each module, and comprehensively evaluates the overall effect of the driving active safety robot control method based on artificial intelligence so as to continuously optimize and improve the system performance.
Compared with the prior art, the invention has the beneficial effects that:
according to the invention, through the alarm level module, different types of alarm levels can be finely divided, and corresponding response modes are set. When facing various driving safety conditions, the system can take different measures according to the severity of the system, and the pertinence and the effectiveness of the alarm are improved. Meanwhile, by the aid of the combined mode of equipment alarming and platform alarming, the fact that relevant personnel can receive accurate information in time is guaranteed, and accordingly quick response is achieved.
And due to the arrangement of the alarm type module, the monitoring content of driving safety is enriched. Through classifying and regularly configuring various common unsafe behaviors, such as call answering alarm, distraction alarm and the like, the adverse behaviors of a driver can be comprehensively detected, and the risk of accidents caused by human factors is greatly reduced. The rule checking function ensures the integrity and usability of configuration and improves the reliability of the system.
The processing type module introduces an AI robot to realize efficient acquisition, transmission and processing of alarm data. The automatic flow assembly function enables the processing process to be more intelligent and efficient, and improves the response speed and processing capacity to driving safety behaviors.
The alarm flow module builds an efficient alarm processing flow by using natural language processing and statistical analysis methods. From the reception of alarm push data to emergency alarm, AI alarm, high frequency alarm and cyclic alarm, continuous supervision and correction of unsafe behavior is ensured until behavior correction, greatly improving the level of assurance of driving safety.
The AI monitoring robot module is used as a control center, so that the safety behavior in the driving process can be accurately identified, and feedback and early warning can be timely performed. By combining an advanced machine learning algorithm and a convolutional neural network, the accuracy of behavior recognition is improved, and powerful technical support is provided for driving safety.
In a word, the driving active safety robot control system based on artificial intelligence effectively improves driving safety level, reduces accident occurrence rate and provides reliable guarantee for life and property safety of drivers and passengers through the synergistic effect of a plurality of modules.
Drawings
FIG. 1 is a schematic block diagram of an artificial intelligence based active safety robot control system for driving a vehicle;
FIG. 2 is a schematic block diagram of an artificial intelligence based driving active safety robot control method according to the present invention;
FIG. 3 is a schematic diagram of the operation steps of an alarm module in the active driving safety robot control system based on artificial intelligence;
FIG. 4 is a schematic diagram of the steps of a warning type module in the driving active safety robot control system based on artificial intelligence;
Fig. 5 is a schematic diagram of steps of a warning flow module in the driving active safety robot control system based on artificial intelligence.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. 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.
In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise" indicate orientations or positional relationships are based on the orientations or positional relationships shown in the drawings, are merely for convenience in describing the present invention and simplify the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and therefore should not be construed as limiting the present invention.
Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more of the described features. In the description of the present invention, the meaning of "a plurality" is two or more, unless explicitly defined otherwise. The terms "mounted," "connected," "coupled," and "connected" are used in a broad sense, and may be, for example, fixedly connected, detachably connected, or integrally connected, mechanically connected, electrically connected, directly connected, or indirectly connected via an intermediate medium, or may be in communication with the interior of two elements. The specific meaning of the terms in the present invention will be understood by those skilled in the art in detail, and the present invention will be further described in detail with reference to the accompanying drawings.
Referring to FIGS. 1-5, an artificial intelligence based active safety robot control system for a vehicle comprises:
the alarm level module is used for setting different types of alarm levels and setting corresponding responses, the alarm levels define the response levels of different alarms, and the different alarm levels can be started in combination with actual practice;
The alarm type module is used for setting different driving alarm types, configuring corresponding rules, setting corresponding alarm levels and alarm attributes (call receiving alarm, distraction alarm, rear approach alarm, smoke alarm, probe shielding alarm and unbelted alarm) according to different alarm types by rule configuration, setting corresponding alarm levels, recording different rule codes as S1, S2 and S3..SN according to the different alarm types, classifying according to S= { S1, S2 and S3..SN, and judging the integrity of the classified configuration by rule verification to ensure that the rules are available;
The processing type module is used for monitoring the generated alarm data, setting different driving safety behavior processing modes, configuring corresponding rules, including alarm data acquisition, data transmission and introduction of an AI robot in processing, receiving alarm information data by the robot during processing, acquiring alarm level configuration through a data acquisition center, and carrying out automatic flow assembly by combining corresponding flow treatment, and sending the alarm data to a data processing center after completion so as to facilitate alarm data processing and execution;
The alarm flow module is used for setting how to treat after the alarm is generated, configuring corresponding execution rules, carrying out dynamic flow execution after receiving the data of the processing type module, carrying out real-time analysis and processing on the acquired data by the flow mainly by using a natural language processing technology and a statistical analysis method, and constructing an intention recognition model by a model construction unit based on a machine learning algorithm and a finite element analysis method, so that the alarm processing flow is convenient to accurately judge, the corresponding treatment can be responded efficiently, the algorithm can also be iterated continuously by settling the data in the driving active safety field, and the complex alarm flow is convenient to recognize and process;
The AI monitoring robot module is a control center for monitoring driving active safety behaviors, is used for setting a behavior monitoring mode, and is used for monitoring after correction, the safety behaviors in the driving can be identified according to different images by dynamically inputting an early warning rule, and are fed back to different users for prompt and alarm, when the occurrence of the safety behaviors is detected or a preset alarm rule is triggered, the early warning information is notified by an automatic short message, and the alarm rule is set and adjusted according to the actual running condition;
the AI knowledge base module is used for setting different active safety knowledge bases, classifying, pushing and early warning, can be continuously perfected based on the existing safety behavior knowledge base, realizes continuous enrichment and perfection of the AI knowledge base through the processes of receiving, processing and feeding back, and is convenient for better forming active safety monitoring;
The group information module manages user group data, identifies user identities and group members in the group, utilizes JSON format and user-defined multi-group information, formulates a time node sequence set according to user personalized pushing requirements to identify a pushing scheme of a target data service, obtains a user personalized pushing scheme set, and carries out short message pushing management of the target data service based on the user personalized pushing scheme.
In the invention, the alarm level module comprises the following operation steps:
Respectively setting a set for the alarm levels, wherein the set is used for carrying out grading response according to the alarm levels and can be formed for the alarm levels;
and setting equipment alarm and platform alarm modes for different alarm levels respectively, realizing equipment alarm mode definition and platform alarm mode definition sets under different alarm levels, combining the equipment alarm mode definition and the platform alarm mode definition sets, and importing the obtained M concentrated information entropy sets into the module as alarm level information sets to finally obtain alarm level data information.
And (3) introducing an alarm level response stability formula: WhereinIndicating the stability of the response of the alarm stage,The total number of alarms is indicated and,Indicating the response accuracy (the number of response accuracy divided by the total number of times) of the ith alarm,Indicating the accuracy of the average response and,Indicating the maximum response accuracy rate and,Representing the minimum response accuracy. The formula evaluates the response stability of the alarm level module by measuring the fluctuation condition of the response accuracy rate and the relation with the maximum response accuracy rate and the minimum response accuracy rate, and ensures that the system can stably respond according to the alarm level under different conditions.
In the invention, the alarm type module comprises the following operation steps:
The method is used for setting different driving alarm types, configuring corresponding rules, setting corresponding alarm levels and alarm attributes (call receiving alarm, distraction alarm, rear approaching alarm, smoking alarm, probe shielding alarm and unbelted safety belt alarm) according to different alarm types, and setting corresponding alarm levels;
And (3) recording configurations of different alarm types as different rule codes, namely S= { S1, S2 and S3..SN }, classifying according to S= { S1, S2 and S3..SN }, and performing integrity judgment on the classified configurations by rule verification to ensure that the rule is available.
Introducing an alarm type configuration complexity formula: WhereinRepresenting the complexity of the alarm type configuration,The number of alarm types is indicated,A weight coefficient representing the ith alarm type (determined based on the importance of the alarm type and frequency of occurrence factors),The complexity of rule configuration (which can be measured by the number of rules and the number of parameters) representing the ith alarm type. The formula comprehensively considers the weights of different alarm types and the rule configuration complexity, and evaluates the configuration complexity of the alarm type module so as to better optimize and manage.
In the invention, the processing type module specifically comprises:
Setting different driving safety behavior processing modes, configuring corresponding rules, including alarm data acquisition, data transmission and introduction of an AI robot in processing, receiving alarm information data by the robot during processing, simultaneously acquiring alarm level configuration through a data acquisition center, and carrying out automatic flow assembly by combining corresponding flow treatment, and sending the alarm data to a data processing center after completion so as to facilitate alarm data processing and execution.
Introducing a treatment efficiency formula: WhereinThe processing efficiency is indicated by the fact that,Indicating the number of alarms that are correctly handled,The total number of alarms is indicated and,Representing the average processing time. The formula comprehensively measures the working efficiency of the processing type module by considering the proportion of the correct processing times to the total alarm times and the reciprocal of the average processing time, so that the system can be ensured to process alarm data rapidly and effectively.
In the invention, the alarm flow module comprises the following steps:
s1.1, receiving alarm push data, namely, transmitting a task to a positioning equipment module through JT808 unified protocol, wherein the positioning equipment module is responsible for receiving and analyzing a task instruction, and ensuring the accuracy and the integrity of task information;
s1.2, pushing related early warning and voice information to related responsible persons through a robot after the receiving equipment pushes an alarm, and pushing different templates according to different types of people;
S1.3, AI alarms, namely after the pushing task is completed, the system emergently sends an emergency alarm voice instruction to alarm equipment through JT808, detects whether pushing alarm is completed or not, and dynamically identifies relevant voices according to alarm information;
S1.4, high-frequency alarming, namely sending an instruction to alarm equipment through a robot, grabbing an active safety behavior correction state, carrying out identification verification on a server, and if the correction is not carried out, starting a high-frequency alarming flow and continuously sending a voice instruction;
S1.5, loop alarm, namely detecting the behavior after high-frequency alarm, if the behavior is not changed, starting the loop alarm, and continuously executing S1.1> S1.2> S1.3> S1.4> S1.5 until the behavior is corrected.
Introducing an alarm process integrity formula: WhereinThe integrity of the alarm flow is indicated,Indicating the number of alarm process steps that were successfully performed,The number of steps in the total alarm flow is indicated,The ratio of the average execution time per step to the maximum allowed execution time is expressed. The formula evaluates the integrity and the execution effect of the alarm flow module by considering the proportion of the number of successfully executed steps to the total number of steps and the rationality of the execution time of each step, so as to ensure that the alarm flow can be executed efficiently and completely.
In the present invention, the AI-monitoring robot module specifically includes:
the data storage unit is used for storing the data uploaded by the active safety behavior alarm data transmission unit and ensuring that the data cannot be lost;
The data analysis unit processes the transmitted data, performs real-time analysis and processing on the acquired data by using a natural language processing technology and a statistical analysis method, the model construction unit constructs an intention recognition model based on a machine learning algorithm and a finite element analysis method, learns and trains data information related to active safety behaviors by adopting a Support Vector Machine (SVM) and an Artificial Neural Network (ANN) machine learning algorithm, establishes a mapping relation between the intention and an entity, optimizes model parameters by combining the finite element analysis method, improves the accuracy and the reliability of the model, and ensures that the prediction accuracy of the model is not lower than 90% on a test set.
In the AI monitoring robot model construction step, vector conversion is carried out on the corresponding grabbing images, and behavior recognition in the images is carried out through a convolutional neural network.
Convolutional neural networks are typically stacked of convolutional layers, pooling layers, and fully-connected layers. The convolution layer utilizes a plurality of different convolution kernels to extract the characteristics of the target and generate a characteristic map, the pooling layer is used for downsampling, combining the characteristics of adjacent characteristic maps and reducing the dimension, and the full-connection layer plays a role in mapping the learned distributed characteristics to a sample marking space.
Introducing a behavior recognition accuracy rate improvement rate formula: WhereinRepresents the improvement rate of the behavior recognition accuracy,The accuracy of behavior recognition after optimization is represented,Representing the accuracy of behavior recognition before optimization. The formula is used for measuring the improvement degree of the behavior recognition accuracy of the AI monitoring robot module before and after optimization, and ensuring that the recognition capability of the system on the driving safety behavior can be continuously improved.
The invention identifies user identity and group members in a group by the group information module, utilizes JSON format and user-defined multi-group information, formulates a time node sequence set according to user personalized pushing requirements to identify a pushing scheme of a target data service, obtains a user personalized pushing scheme set, and carries out short message pushing management of the target data service based on the user personalized pushing scheme.
And introducing a group information push accuracy formula: WhereinRepresents the group information push accuracy rate,Indicating the number of information that was correctly pushed to the target user,Indicating the total number of pushes of information. The formula is used for evaluating the accuracy of the group information module for pushing the user information, and ensuring that the system can accurately push the related information to the target user.
The invention discloses a driving active safety robot control method based on artificial intelligence, which comprises the following steps:
setting alarm levels, namely setting different types of alarm levels to form an alarm level set, defining equipment alarm and platform alarm modes for the different alarm levels, and importing alarm level information sets into a system;
Setting different driving alarm types, configuring alarm levels and alarm attribute rules according to the alarm types, classifying the different alarm types and checking the rules;
setting different driving safety behavior processing modes, configuring corresponding rules including alarm data acquisition and transmission and AI robot flow assembly, and sending the processed data to a data processing center;
The alarm process execution step is that the JT808 unified protocol is used for receiving alarm push data and carrying out emergency alarm push, AI alarm, high-frequency alarm and cyclic alarm until the safety behavior is corrected;
The AI monitoring step comprises the steps of monitoring driving safety behaviors by using an AI monitoring robot module, including alarm data storage and analysis, performing behavior recognition by using a convolutional neural network, constructing an intention recognition model, and improving the accuracy of behavior recognition;
Setting different active safety knowledge bases, and continuously enriching and perfecting the AI knowledge bases through receiving, processing and feedback processes to realize classified pushing early warning;
and a group information management step of managing user group data, identifying group user identities and members, and performing short message pushing management according to user personalized pushing requirements by utilizing JSON format and user-defined multi-group information.
The method is introduced into a comprehensive effect evaluation formula:
WhereinThe method is represented by a comprehensive effect evaluation value,Indicating the stability of the response of the alarm stage,Representing the complexity of the alarm type configuration,The processing efficiency is indicated by the fact that,The integrity of the alarm flow is indicated,Represents the improvement rate of the behavior recognition accuracy,Represents the group information push accuracy rate,、、、、、The weight coefficient is set according to the importance degree of different indexes. The formula comprehensively considers key indexes of each module, and comprehensively evaluates the overall effect of the driving active safety robot control method based on artificial intelligence so as to continuously optimize and improve the system performance.
The present invention is not limited to the above-mentioned embodiments, and any person skilled in the art, based on the technical solution of the present invention and the inventive concept thereof, can be replaced or changed within the scope of the present invention.