An optimization control method and system for intelligent industry

Abstract

The application relates to the field of industrial safety control, and discloses an intelligent industrial optimization control method and system. The method and system realize the timely discovery and tracking of moving objects in a region by adopting a global real-time monitoring mode, perform fitting calculation on the motion state of the moving objects based on monitoring data to predict the motion path of the moving objects in a future time period, judge the possibility and time retention of collision with a pipeline or falling off the pipeline, and then can perform pre-shutdown on the pipeline equipment, effectively avoiding or reducing the secondary injury of personnel and equipment caused by the personnel being involved after the personnel operating emergency stop stagnation due to the falling of the articles and personnel off the pipeline equipment, and effectively reducing the production loss risk.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of industrial safety control, and particularly relates to an intelligent industrial optimization control method and system. BACKGROUND

[0002] In industrial production, when a safety accident such as falling of an article or a person occurs (especially for structures with advancing or rotating capabilities such as conveyors, winding devices, etc.), not only will the equipment be damaged, causing economic losses, but the falling person may even be injured, becoming a serious production safety accident. Therefore, in order to deal with sudden safety accidents, industrial production equipment is usually provided with an emergency stop button for controlling the equipment to stop immediately.

[0003] However, with the increasing intelligence of current assembly line equipment, the demand for manual work is gradually decreasing. For example, only one or two workers may be assigned to a device, which may result in the device not being discovered and the emergency stop measure not being performed in time by other workers in the event of an accident, causing the accident to expand. Even if the workers discover and control the emergency stop in time, the manual emergency stop method has a certain lag and has a large safety risk. SUMMARY

[0004] The present application aims to provide an intelligent industrial optimization control method and system to solve the problems in the background.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0006] An intelligent industrial optimization control method, comprising:

[0007] Real-time monitoring of a risk monitoring area, object motion state recognition based on real-time monitoring feedback image data, and continuous tracking monitoring of a moving object, wherein the moving object is used to represent a non-static object;

[0008] Calculation fitting of the motion state of the moving object based on continuous tracking monitoring data, acquisition of motion characteristic parameters thereof, motion prediction based on the motion characteristic parameters, and generation of a predicted trajectory of the moving object, wherein the motion characteristic parameters include a motion direction and a multi-direction motion component;

[0009] Collision determination of the moving object according to the predicted trajectory, wherein if the determination result represents that a collision with a production device will occur, the moving object is marked as a collision high-risk object;

[0010] An emergency triggering range is set, if the collision high-risk object enters the emergency triggering range, an emergency control signal is generated based on the instantaneous speed of the collision high-risk object and the flow line equipment is controlled to stop, the emergency control signal corresponds to a parking limit time, the parking limit time is inversely proportional to the instantaneous speed, and the parking limit time is used to represent the maximum available time for the flow line to complete the parking action.

[0011] As a further scheme of the present application, the step of distinguishing the subjective motion object is further included, and specifically includes:

[0012] Based on the real-time monitoring data, image features of the motion object are selected to obtain corresponding image recognition features, and the object category of the motion object is obtained by searching according to the image recognition features, wherein the object category includes biological objects and article objects.

[0013] If the object category represents a biological object, a biological marker is added to the motion object, and the position information and characteristic motion information of the motion object are recorded in real time.

[0014] As a further scheme of the present application, the step of generating the predicted trajectory of the motion object based on the motion characteristic parameters further includes:

[0015] When the motion object is a biological object, an imbalance determination is triggered, and the characteristic motion data of the motion object is disassembled into characteristic objects, wherein the characteristic objects represent different nodes of different limb parts of the biological object, including a trunk and limbs, and the trunk and limbs each include a plurality of characteristic nodes.

[0016] The motion state of the plurality of characteristic objects of the motion object is calculated to obtain motion data of the corresponding characteristic nodes, wherein the motion data includes a motion direction, a motion speed, a motion acceleration, a trunk angular velocity, and a spatial correlation between nodes.

[0017] The motion data is evaluated by a big data imbalance model, if the evaluation result represents an object imbalance, a fall warning is generated and a motion characteristic parameter is established based on an imbalance direction, and the evaluation result includes a balance state of the biological object and imbalance direction information in an imbalance state.

[0018] As a further scheme of the present application, the step of monitoring and returning is further included:

[0019] A plurality of return hierarchical distances are set based on the boundary range of the flow line equipment, a plurality of hierarchical intervals are established based on the return hierarchical distances, and the minimum range of the hierarchical intervals is greater than the emergency triggering range.

[0020] According to the hierarchical interval setting, a real-time monitoring return rate is set, and distances of different hierarchical intervals and boundary ranges are inversely proportional to the real-time monitoring return rate, and the real-time monitoring return rate is used to represent the range time and the number of monitoring data acquisition times of the moving object.

[0021] As a further scheme of the present application, an emergency prevention step is further included, specifically comprising:

[0022] If the moving object enters the emergency trigger range and the moving object does not include a collision high-risk object mark, a prevention deceleration signal is generated to control the pipeline equipment to run at a reduced speed.

[0023] The monitoring return rate of the moving object is increased to efficiently judge the motion state change of the moving object, and the collision bouncing motion of the moving object is predicted based on the three-dimensional distribution of the surrounding equipment structure to judge the possibility of the moving object colliding with the environment structure to change the motion state and then colliding with the production equipment.

[0024] The embodiment of the present application aims to provide an intelligent industrial optimization control system, comprising:

[0025] An object monitoring recording module is used to perform real-time monitoring on the risk monitoring area, perform object motion state recognition based on image data fed back by real-time monitoring, and continuously track and monitor the moving object, the moving object being used to represent a non-static object;

[0026] A motion fitting prediction module is used to calculate and fit the motion state of the moving object based on continuous tracking and monitoring data, obtain motion characteristic parameters thereof, and perform motion prediction based on the motion characteristic parameters to generate a predicted trajectory of the moving object, the motion characteristic parameters including a motion direction and motion components in multiple directions.

[0027] An object collision prediction module is used to perform collision determination on the moving object according to the predicted trajectory, and if the determination result represents that the moving object will collide with the production equipment, the moving object is marked as a collision high-risk object.

[0028] An emergency trigger control module is used to set an emergency trigger range, and if the collision high-risk object enters the emergency trigger range, an emergency control signal is generated based on the instantaneous speed of the collision high-risk object to control the pipeline equipment to stop, the emergency control signal corresponding to include a stop limit time, the stop limit time being inversely proportional to the instantaneous speed, and the stop limit time being used to represent the maximum available time length for the pipeline to complete a stop action.

[0029] As a further scheme of the present application, a subjective object distinguishing module is further included, comprising:

[0030] The subjective feature distinguishing unit is configured to perform image feature selection on the moving object based on the real-time monitoring data, to obtain corresponding image recognition features, and to perform retrieval according to the image recognition features, to obtain an object category of the moving object, wherein the object category includes biological objects and article objects.

[0031] The biological object marking unit is configured to perform biological marking on the moving object if the object category represents a biological object, and to record position information and characteristic motion information of the moving object in real time.

[0032] As a further scheme of the present application, the motion fitting prediction module comprises:

[0033] The imbalance determination triggering unit is configured to trigger imbalance determination when the moving object is a biological object, to perform feature object decomposition on the characteristic motion data of the moving object, and to use the feature objects to represent different nodes of different limb parts of the biological object, including a trunk and limbs, wherein the trunk and the limbs each include a plurality of feature nodes.

[0034] The determination data acquisition unit is configured to perform motion state calculation on a plurality of feature objects of the moving object, to obtain motion data of corresponding feature nodes, and to include motion direction, motion speed, motion acceleration, trunk angular velocity, and spatial correlation between nodes in the motion data.

[0035] The imbalance determination execution unit is configured to evaluate the motion data by using a big data imbalance model, to generate a fall warning if the evaluation result represents object imbalance, and to establish motion feature parameters based on an imbalance direction, wherein the evaluation result includes a balance state of the biological object and imbalance direction information in an imbalance state.

[0036] As a further scheme of the present application, the return level demarcation module comprises:

[0037] The return grading unit is configured to set a plurality of return grading distances based on a boundary range of the flow device, to establish a plurality of grading intervals based on the return grading distances, and to set a minimum range of the grading intervals to be greater than the emergency triggering range.

[0038] The return calibration unit is configured to set a real-time monitoring return rate according to the grading intervals, to make the distance of different grading intervals and boundary ranges be inversely proportional to the real-time monitoring return rate, and to use the real-time monitoring return rate to represent the number of times of monitoring data acquisition of the moving object within a range time.

[0039] As a further scheme of the present application, the emergency prevention module comprises:

[0040] A prevention triggering unit is configured to generate a prevention deceleration signal to control the pipeline equipment to operate at a low speed if the moving object enters an emergency triggering range and the moving object does not include a collision high-risk object mark.

[0041] A prevention tracing unit is configured to increase the monitoring return rate of the moving object, efficiently judge the change of the moving state of the moving object, and predict the collision bounce movement of the moving object based on the three-dimensional distribution of the surrounding equipment structure, to judge the possibility of the moving object colliding with the environment structure again to produce the equipment after the change of the moving state.

[0042] Compared with the prior art, the present application has the beneficial effects that: by adopting the global real-time monitoring mode, the moving object in the region is discovered and tracked in time, and the motion state of the moving object is fitted and calculated based on the monitoring data to predict the motion path of the moving object in the future time period, to judge the possibility and time retention of the collision with the pipeline or the falling of the pipeline, so that the pipeline equipment can be pre-stopped, the secondary injury of personnel and equipment caused by the personnel operation after the sudden stop of the personnel falling into the pipeline equipment can be effectively avoided or reduced, and the production loss risk can be effectively reduced. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 It is a flowchart of an intelligent industrial optimization control method.

[0044] Figure 2 It is a flowchart of an imbalance direction prediction step in an intelligent industrial optimization control method.

[0045] Figure 3 It is a composition diagram of an intelligent industrial optimization control system. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application is further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application, and are not used to limit the present application.

[0047] The specific implementation mode of the present application is described in detail below in combination with specific examples.

[0048] As Figure 1 The intelligent industrial optimization control method provided by an embodiment of the present application includes the following steps:

[0049] S10, real-time monitoring is performed on a risk monitoring region, object motion state recognition is performed based on image data fed back by real-time monitoring, and continuous tracking monitoring is performed on a moving object, the moving object being used to represent a non-static object.

[0050] S20, performing calculation fitting on the motion state of the motion object based on the continuous tracking monitoring data, obtaining motion characteristic parameters of the motion object, and performing motion prediction based on the motion characteristic parameters to generate a predicted trajectory of the motion object, the motion characteristic parameters including a motion direction and a multi-direction motion component;

[0051] S30, performing collision determination on the motion object according to the predicted trajectory, and if the determination result represents that a collision with the production equipment will occur, marking the motion object as a collision high-risk object;

[0052] S40, setting an emergency triggering range, if the collision high-risk object enters the emergency triggering range, generating an emergency control signal based on an instantaneous speed of the collision high-risk object and controlling the flow line equipment to stop, the emergency control signal corresponding to including a stop limit time, the stop limit time being inversely proportional to the instantaneous speed, and the stop limit time being used to represent a maximum available duration for the flow line to complete a stop action.

[0053] In this embodiment, an intelligent industrial optimization control method is given. The global real-time monitoring method is adopted to realize the timely discovery and tracking of the moving objects in the region, and the motion state of the moving object is fitted and calculated based on the monitoring data to predict the motion path of the moving object in the future time period, to judge the possibility and time retention of collision with the assembly line or falling off the assembly line, and then the assembly line equipment can be pre-stopped to effectively avoid or reduce the secondary injury of personnel and equipment caused by the personnel operation emergency stop after the falling of the goods and personnel into the assembly line equipment, and effectively reduce the production loss risk. In the prior art, in factory production, especially in some scenes involving the use of assembly line production equipment, there will be a conveying process of production goods between adjacent devices or adjacent process technologies of the same device (or rotating type devices). However, this process often easily causes safety production accidents in actual production, such as accidental involvement of personnel, accidental involvement of other goods and equipment, etc. Because the related equipment in the prior art usually adopts manual emergency stop when encountering such a situation, that is, the workers perform emergency parking through the emergency stop button set by the machine. However, such a processing method will inevitably cause a certain degree of involvement due to the delay of manual reaction. If the involved object is goods, it may cause damage to the equipment, resulting in a long maintenance period and causing great economic losses. If the involved object is personnel, it will not only cause economic losses, but also cause personnel casualties and have a bad impact. The specific implementation process for solving this problem in this embodiment is as follows: when a moving object (such as a falling or flying object, a falling personnel, etc.) is detected, the motion state of the object is calculated based on the data obtained by the monitoring, and then the motion of the object is simulated according to the motion physical model to predict the motion trajectory of the object in the future period of time. According to the predicted trajectory, it can be judged whether the object will collide with the corresponding region of the assembly line equipment or fall into the equipment, and then the parking control of the equipment can be realized in a small period of time in advance, effectively preventing the involvement of the object in the equipment due to the movement of the equipment when the object falls into the equipment. Further, the parking time limit is given in the embodiment. The purpose of this setting is to make the parking process a process step, that is, to complete the transition in a period of time, rather than to stop immediately to 0 of the equipment running speed, which can protect the equipment to a certain extent. The time is determined according to the speed and distance of the moving object, that is, the predicted time of contact between the moving object and the equipment.

[0054] As another preferred embodiment of the present application, it further includes a subjective moving object distinguishing step, specifically including:

[0055] The image feature of the moving object is selected based on the real-time monitoring data to obtain corresponding image recognition features, and the object category of the moving object is obtained by searching according to the image recognition features, wherein the object category includes biological objects and article objects.

[0056] If the object category represents a biological object, a biological marker is added to the moving object, and the position information and characteristic motion information of the moving object are recorded in real time.

[0057] Further, as shown in Figure 2 As another preferred embodiment of the present application, the step of generating the predicted trajectory of the moving object based on the motion characteristic parameters further includes:

[0058] S51, when the moving object is a biological object, triggering imbalance determination, and the characteristic motion data of the moving object is decomposed into feature objects, wherein the feature objects represent different nodes of different limb parts of the biological object, including the trunk and limbs, and the trunk and limbs each include a plurality of feature nodes;

[0059] S52, calculating the motion state of the plurality of feature objects of the moving object to obtain the motion data of the corresponding feature nodes, wherein the motion data includes motion direction, motion speed, motion acceleration, trunk angular velocity, and spatial correlation between nodes;

[0060] S53, evaluating the motion data by a big data imbalance model, and if the evaluation result represents object imbalance, generating a fall warning and establishing motion characteristic parameters based on the imbalance direction, wherein the evaluation result includes the balance state of the biological object and the imbalance direction information in the imbalance state.

[0061] In the actual execution process of this embodiment, the human body is different from the still object. The motion of the still object can be simply predicted according to the physical law, but the motion mode of the human body is subjective. The still object flying towards the assembly line equipment can be simply trajectory predicted by motion calculation, and the risk is determined according to whether the trajectory overlaps with the equipment. However, the motion of the staff in the factory area is subjective and unpredictable, so in daily work, the motion mode of the staff walking towards the assembly line cannot be simply determined as a risk condition, therefore, the process of judging the human body posture is supplemented here, that is, whether the human body is imbalance is judged by the posture data of the human body, such as the acceleration of the hand and the angular velocity change of the skeleton, when the staff is imbalance, it means that he may fall down, or even fall down on the production equipment, and after the human body is imbalance, its motion state is difficult to be effectively controlled by the human body, therefore, the motion prediction can be performed according to the current motion state to determine whether it will overlap with the equipment and cause an accident.

[0062] As another preferred embodiment of the present application, the monitoring return demarcation step is further included.

[0063] The multi-return grading distance is set based on the boundary range of the flow equipment, and a plurality of grading intervals are established based on the return grading distance, the minimum range of the grading intervals being greater than the emergency trigger range;

[0064] The real-time monitoring return rate is set according to the grading intervals, the distance of different grading intervals and boundary ranges being inversely proportional to the real-time monitoring return rate, and the real-time monitoring return rate being used to represent the range-time monitoring data acquisition times of the moving object.

[0065] In the embodiment, the grading intervals are defined, different data monitoring collection in different grading intervals adopts different data return frequencies (i.e., the closer to the flow line equipment, the higher the data return frequency), through such a management mode, the data return bandwidth demand can be effectively reduced to be reasonably allocated to the risk situation, so as to ensure that there is sufficient data bandwidth for data acquisition in the risk situation, and the energy consumption can be effectively reduced.

[0066] As another preferred embodiment of the present application, the emergency prevention step is further included, and specifically includes:

[0067] If the moving object enters the emergency trigger range and the moving object does not contain the collision high-risk object mark, a prevention deceleration signal is generated to control the flow line equipment to run at a reduced speed.

[0068] The monitoring return rate of the moving object is increased to efficiently judge the motion state change of the moving object, and the collision bouncing motion of the moving object is predicted based on the three-dimensional distribution of the surrounding equipment structure to judge the possibility of the moving object colliding with the environment structure to change the motion state and then colliding with the production equipment.

[0069] In the embodiment, for the emergency trigger range, in addition to triggering the shutdown, the deceleration process can also be triggered, so as to avoid the motion state change of the object judged as not colliding after entering the range, for example, the object judged as not colliding enters the emergency trigger range, the motion state is changed due to collision with other objects or structures, and at this time, because the distance to the equipment is very close, the system has a short judgment time, and therefore, the shutdown may not be timely, therefore, by reducing the vehicle speed, more time is obtained for the subsequent motion change to obtain a faster parking response time.

[0070] As shown in Figure 3 The present application also provides an intelligent industrial optimization control system, which includes:

[0071] The object monitoring recording module 100 is configured to monitor a risk monitoring area in real time, identify a motion state of an object based on image data fed back in real time, and continuously track and monitor the moving object, wherein the moving object is used to represent a non-static object;

[0072] The motion fitting prediction module 200 is configured to calculate and fit the motion state of the moving object based on the continuous tracking and monitoring data, obtain motion characteristic parameters of the moving object, and perform motion prediction based on the motion characteristic parameters to generate a predicted trajectory of the moving object, wherein the motion characteristic parameters include a motion direction and motion components in multiple directions.

[0073] The object collision prediction module 300 is configured to determine collision of the moving object according to the predicted trajectory, and if the determination result indicates that the moving object will collide with the production equipment, mark the moving object as a collision high-risk object.

[0074] The emergency triggering control module 400 is configured to set an emergency triggering range, if the collision high-risk object enters the emergency triggering range, generate an emergency control signal based on an instantaneous speed of the collision high-risk object and control the flow line equipment to stop, wherein the emergency control signal corresponds to a stop limit time, the stop limit time is inversely proportional to the instantaneous speed, and the stop limit time is used to represent a maximum available time length for the flow line to complete a stopping action.

[0075] As another preferred embodiment of the present application, the subjective object distinguishing module comprises:

[0076] The subjective feature distinguishing unit is configured to select image features of the moving object based on the real-time monitoring data to obtain corresponding image recognition features, and search according to the image recognition features to obtain an object category of the moving object, wherein the object category comprises a biological object and an article object.

[0077] The biological object marking unit is configured to mark the moving object as a biological object if the object category represents a biological object, and track and record position information and characteristic motion information of the moving object in real time.

[0078] As another preferred embodiment of the present application, the motion fitting prediction module comprises:

[0079] The imbalance determination triggering unit is configured to trigger imbalance determination when the moving object is a biological object, and decompose characteristic objects of the characteristic motion data of the moving object, wherein the characteristic objects are used to represent different nodes of different limb parts of the biological object, including a trunk and limbs, and the trunk and the limbs each comprise a plurality of characteristic nodes.

[0080] The determination data acquisition unit is configured to perform motion state calculation on the plurality of feature objects of the moving object, and acquire motion data of the corresponding feature nodes, wherein the motion data comprises motion direction, motion speed, motion acceleration, trunk angular velocity, and spatial correlation between nodes.

[0081] The imbalance determination execution unit is configured to evaluate the motion data by using a big data imbalance model, and if the evaluation result represents that the object is imbalanced, generate a fall warning and establish motion feature parameters based on the imbalance direction, wherein the evaluation result comprises the balance state of the biological object and the imbalance direction information in the imbalance state.

[0082] As another preferred embodiment of the present application, the return level demarcation module comprises:

[0083] The return grading unit is configured to set a plurality of return grading distances based on the boundary range of the flow equipment, and establish a plurality of grading intervals based on the return grading distances, wherein the minimum range of the grading intervals is greater than the emergency trigger range.

[0084] The return calibration unit is configured to set a real-time monitoring return rate according to the grading intervals, wherein the distance of different grading intervals and boundary ranges is inversely proportional to the real-time monitoring return rate, and the real-time monitoring return rate is used to represent the range time and the number of times of monitoring data acquisition of the moving object.

[0085] As another preferred embodiment of the present application, the emergency prevention module comprises:

[0086] The prevention trigger unit is configured to generate a prevention deceleration signal to control the flow line equipment to operate at a reduced speed if the moving object enters the emergency trigger range and the moving object does not contain a collision high-risk object label.

[0087] The prevention tracing unit is configured to increase the monitoring return rate of the moving object to efficiently judge the motion state change of the moving object, and predict the collision and rebound motion of the moving object based on the three-dimensional distribution of the surrounding equipment structure, to judge the possibility of the moving object colliding with the environment structure to change the motion state and then colliding with the production equipment.

[0088] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The program can be stored in a non-volatile computer readable storage medium, and when the program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, storage, databases, or other media in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in many forms such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0089] Other embodiments of the present disclosure will be apparent to those skilled in the art with the accomplishment of the present disclosure as reflected in the specification and embodiments. The present application is intended to cover any variations, uses, or adaptive changes of the present disclosure that follow the general principles of the present disclosure and include common knowledge or conventional technical means in the art that are not disclosed by the present disclosure. The specification and embodiments are only regarded as exemplary, and the true scope and spirit of the present disclosure are indicated by the claims.

[0090] It should be understood that the present disclosure is not limited to the precise structures described above and shown in the drawings and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is only limited by the appended claims.

Claims

1. An optimal control method for smart industry, characterized by, Comprise: Real-time monitoring of the risk monitoring area, object motion state recognition based on real-time monitoring feedback image data, and continuous tracking monitoring of the moving object, which is used to represent a non-static object; Based on the continuous tracking monitoring data, the motion state of the moving object is calculated and fitted to obtain its motion characteristic parameters, and motion prediction is performed based on the motion characteristic parameters to generate the predicted trajectory of the moving object, the motion characteristic parameters including the motion direction and the motion component in multiple directions; According to the predicted trajectory, the collision of the moving object is determined, and if the determination result represents that the collision with the production equipment will occur, the moving object is marked as a collision high-risk object; Set an emergency trigger range, if the collision high-risk object enters the emergency trigger range, generate an emergency control signal based on the instantaneous speed of the collision high-risk object and control the pipeline equipment to stop, the emergency control signal corresponding to the stop limit time, the stop limit time being inversely proportional to the instantaneous speed, the stop limit time representing the maximum available time for the pipeline to complete the stopping action; The step of generating the predicted trajectory of the moving object based on the motion characteristic parameters also includes: When the moving object is a biological object, trigger the imbalance determination, and perform feature object decomposition on the feature motion data of the moving object, the feature object representing different nodes of different limb parts of the biological object, including the trunk and the limbs, and the trunk and the limbs each including multiple feature nodes; Perform motion state calculation on multiple feature objects of the moving object to obtain motion data of corresponding feature nodes, the motion data including motion direction, motion rate, motion acceleration, trunk angular velocity, and spatial correlation between nodes; Evaluate the motion data through a big data imbalance model, and if the evaluation result represents that the object is imbalanced, generate a fall warning and establish motion characteristic parameters based on the imbalance direction, the evaluation result including the balance state of the biological object and the imbalance direction information in the imbalance state.

2. The optimal control method for smart industry according to claim 1, wherein, It also includes a step of distinguishing subjective motion objects, specifically including: Based on real-time monitoring data, select image features of the moving object to obtain corresponding image recognition features, and retrieve according to the image recognition features to obtain the object category of the moving object, the object category including biological objects and article objects; If the object category represents a biological object, the moving object is labeled as a biological object, and its position information and feature motion information are recorded in real time.

3. The optimal control method for smart industry according to claim 1, wherein, It also includes a monitoring reward setting step: Based on the boundary range of the flow equipment, set multiple reward classification distances, and based on the reward classification distances, establish multiple classification intervals, the minimum range of the classification interval being greater than the emergency trigger range; According to the classification interval, set a real-time monitoring reward rate, the distance of different classification intervals and boundary ranges being inversely proportional to the real-time monitoring reward rate, the real-time monitoring reward rate representing the number of times of monitoring data acquisition of the moving object in a certain range.

4. The optimal control method for smart industry according to claim 3, wherein, It also includes an emergency prevention step, specifically including: If the moving object enters the emergency trigger range and the moving object does not contain the collision high-risk object mark, a preventive deceleration signal is generated to control the pipeline equipment to operate at a reduced speed; The monitoring return rate of the moving object is increased to efficiently determine the change in the motion state of the moving object, and the collision bouncing motion of the moving object is predicted based on the three-dimensional distribution of the surrounding equipment structure to determine the possibility of the moving object colliding with the environment structure to change the motion state and then colliding with the production equipment.

5. An optimized control system for smart industry, characterized in that, Comprise: The object monitoring recording module is used for real-time monitoring of the risk monitoring area, object motion state recognition based on real-time monitoring feedback image data, and continuous tracking monitoring of the moving object, the moving object being used to represent a non-static object; The motion fitting prediction module is used for calculating and fitting the motion state of the moving object based on continuous tracking monitoring data, obtaining motion characteristic parameters thereof, and performing motion prediction based on the motion characteristic parameters to generate a predicted trajectory of the moving object, the motion characteristic parameters including a motion direction and a plurality of directional motion components; The object collision prediction module is used for collision determination of the moving object according to the predicted trajectory, and if the determination result represents that a collision with the production equipment will occur, the moving object is marked as a collision high-risk object; The emergency trigger control module is used for setting an emergency trigger range, and if the collision high-risk object enters the emergency trigger range, an emergency control signal is generated based on the instantaneous speed of the collision high-risk object and the pipeline equipment is stopped, the emergency control signal corresponding to including a parking restriction time, the parking restriction time being inversely proportional to the instantaneous speed, and the parking restriction time being used to represent the maximum available time length for the pipeline to complete the parking action; The motion fitting prediction module comprises: The imbalance determination trigger unit is used for triggering imbalance determination when the moving object is a biological object, and feature object decomposition is performed on the feature motion data of the moving object, the feature objects being used to represent different nodes of different limb parts of the biological object, including a trunk and limbs, and the trunk and limbs each including a plurality of feature nodes; The determination data acquisition unit is used for performing motion state calculation on a plurality of feature objects of the moving object to obtain motion data of the corresponding feature nodes, the motion data including a motion direction, a motion rate, a motion acceleration, a trunk angular velocity, and a spatial correlation between nodes; The imbalance determination execution unit is used for evaluating the motion data through a big data imbalance model, and if the evaluation result represents that the object is imbalanced, a fall warning is generated and motion characteristic parameters are established based on the imbalance direction, the evaluation result including a balance state of the biological object and imbalance direction information in the imbalance state.

6. The optimal control system of smart industry according to claim 5, wherein, Further comprising a subjective object distinguishing module, comprising: The subjective feature distinguishing unit is used for selecting image features of the moving object based on real-time monitoring data to obtain corresponding image recognition features, and retrieving according to the image recognition features to obtain an object category of the moving object, the object category including biological objects and article objects. The biological object marking unit is configured to mark a biological object and record position information and characteristic motion information of the moving object in real time if the object category is represented by a biological object.

7. The optimal control system of smart industry according to claim 5, characterized in that, The return level setting module comprises: The return grading unit is configured to set a plurality of return grading distances based on the boundary range of the flow equipment, and establish a plurality of grading intervals based on the return grading distances, wherein the minimum range of the grading intervals is greater than the emergency triggering range; The return setting unit is configured to set a real-time monitoring return rate according to the grading intervals, wherein the distance of different grading intervals and boundary ranges is inversely proportional to the real-time monitoring return rate, and the real-time monitoring return rate is used to represent the range time and the number of times of monitoring data acquisition of the moving object.

8. The optimal control system of smart industry according to claim 7, characterized in that, The emergency prevention module comprises: The prevention triggering unit is configured to generate a prevention deceleration signal to control the flow line equipment to operate at a low speed if the moving object enters the emergency triggering range and the moving object does not contain a collision high-risk object mark. The prevention tracing unit is configured to increase the monitoring return rate of the moving object to efficiently judge the motion state change of the moving object, and predict the collision bouncing motion of the moving object based on the three-dimensional distribution of the surrounding equipment structure to judge the possibility of the moving object colliding with the environment structure to change the motion state and then colliding with the production equipment.

Images