An intelligent debugging and control method and system for AGV vehicles used in unmanned warehousing

Through the remote monitoring platform and warehousing area topology map model, combined with data acquisition and deep learning model, the accuracy of AGV trolley power consumption and status monitoring in complex paths and obstacles is solved, and the operation stability and reliability of the equipment are improved. sex.

CN118519406BActive Publication Date: 2024-11-05TIANJIN MASTER LOGISTICS EQUIP CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410969379.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-19
Publication Date
2024-11-05
Estimated Expiration
2044-07-19

AI Technical Summary

Technical Problem

During the warehousing and distribution process, AGV trolleys face complex paths, obstacles and narrow channels, resulting in an increase in power consumption. Traditional methods cannot accurately estimate power consumption and monitor the status in real time, affecting the reliability and stability of the equipment.

Method used

Build a remote monitoring platform and topology map model of warehousing area, filter key AGV trolleys through data acquisition, simulation testing and deep learning models, monitor operating status in real time, estimate power consumption, and monitor potential faults through fault matrix sets to improve power consumption pre-premise Estimate accuracy and equipment stability.

Benefits of technology

The accuracy of AGV car's power consumption estimate and the reliability of equipment status monitoring are achieved, and the operation stability and reliability of AGV car in complex paths and obstacles is improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118519406B_ABST
    Figure CN118519406B_ABST
Patent Text Reader

Abstract

An intelligent debugging and control method and system for an AGV car for unmanned warehousing, relating to the technical field of AGV cars, constructing a remote monitoring platform to collect the operating data of the AGV car; constructing a storage area topological map model, extracting the key data of the storage order and the operating data of each AGV car, screening out the key AGV car according to the positional relationship between each AGV car and the storage order in the storage area topological map model and the matching relationship between the operating data of the AGV car and the key data; performing a simulation test on the AGV car in advance, obtaining a fault matrix set of the key AGV car according to the simulation test result, monitoring the operating data time series sequence of the key AGV car in the storage order distribution process in real time, monitoring the operating status of the key AGV car according to the operating data time series sequence and the fault matrix set, and significantly improving the reliability and stability of the AGV car storage order distribution process.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of AGV trolleys, and in particular to an intelligent debugging and control method of an AGV trolley for unmanned warehousing. Background Art

[0002] The prior art CN116300775A "A smart logistics AGV trolley control system" includes a monitoring module, which includes a speed sensor and a timer; an acquisition module to obtain the current speed values ​​of several AGV trolleys, the current running time of the arrival point, the number of arrivals, and the total number of times the materials are transported; a calculation module to calculate the speed difference between the current speed value and the minimum speed value, the time difference between the current running time and the standard running time, the time ratio between the current running time and the standard running time, and the accuracy of the AGV trolley arriving at the point; a central control module to determine whether to adjust the AGV trolley according to the running status of the AGV trolley, determine the reason for the adjustment of the AGV trolley, and determine whether the AGV trolley can execute this task route; an early warning module to issue an early warning to the management personnel when the AGV trolley cannot execute this task route; and improve the transportation efficiency of materials.

[0003] The prior art CN113110456B "A method for AGV trolley stable driving based on artificial intelligence" obtains the path image of the trolley under the front view angle; divides the magnetic circuit area in the path image and transforms the front view angle into the top view angle to obtain the top view magnetic circuit map; determines whether there is an abnormal magnetic circuit through the image information and magnetic induction signal in the top view magnetic circuit map; obtains the discontinuous curvature turning point in the top view magnetic circuit map; when the discontinuous curvature turning point moves to the warning area in the top view magnetic circuit convex, adjusts the feedback of the magnetic induction signal through the turning degree of the path; obtains the characteristic area of ​​the discontinuous curvature turning point in the path image, and performs feature matching in the subsequent collected path image. If the matching fails, it indicates that the turning is over. The present invention controls and adjusts the operation of the trolley according to the turning degree by detecting the abnormal magnetic circuit, thereby ensuring the stability of the trolley during operation.

[0004] In the prior art, when the complexity of the warehouse distribution trajectory is high, for example: multiple turns: there are multiple consecutive turns in the path, such as shuttling between shelves inside the warehouse, which requires frequent changes of direction; obstacle crossing: there are obstacles that need to be crossed on the path, such as stacked goods or other AGVs; narrow passages: the path passes through narrow passages or crowded areas, and the AGV needs to operate carefully to avoid collisions; these situations will increase the complexity of autonomous path planning and autonomous dynamic scheduling during the driving process of the AGV, resulting in increased power consumption of the AGV. In the traditional AGV delivery process, the power consumption value of the AGV is often determined based on the delivery distance, and the accuracy of the AGV power consumption estimation and the accuracy of the real-time status monitoring of the AGV cannot be guaranteed. Summary of the invention

[0005] In order to solve the above technical problems, the purpose of the present invention is to provide an intelligent debugging and control method for an AGV car for unmanned warehousing, comprising the following steps:

[0006] Step s1: Build a remote monitoring platform to collect the operation data of the AGV;

[0007] Step s2: Build a storage area topology map model. When the remote monitoring platform receives a storage order, extract the key data of the storage order and the operation data of each AGV. According to the positional relationship between each AGV and the storage order in the storage area topology map model and the matching relationship between the operation data of the AGV and the key data, select the key AGV.

[0008] Step s3: Perform simulation tests on the AGV trolley in advance, obtain the fault matrix set of the key AGV trolley based on the simulation test results, monitor the operating data time series sequence of the key AGV trolley in the warehouse order delivery process in real time, and monitor the operating status of the key AGV trolley based on the operating data time series sequence and the fault matrix set.

[0009] Furthermore, the process of building a remote monitoring platform and collecting the operating data of the AGV includes:

[0010] A remote monitoring platform is constructed, and the remote monitoring platform is connected to a number of data transmission nodes. The data transmission nodes are deployed on the AGV car. The data transmission nodes are used to collect the operating data of the AGV car and mark the collection time and set the collection cycle. The operating data includes the driving position, driving speed, remaining power and remaining load.

[0011] Furthermore, a storage area topology map model is constructed. When the remote monitoring platform receives a storage order, the key data of the storage order and the operation data of each AGV are extracted. According to the positional relationship between each AGV and the storage order in the storage area topology map model and the matching relationship between the operation data of the AGV and the key data, the process of screening out the key AGV includes:

[0012] According to the actual situation, build a topological map model of the storage area;

[0013] The key data of the warehouse order includes the weight of the goods, the order location and the warehouse distribution track. At the same time, the driving position, remaining power and remaining load of each AGV are displayed in the warehouse area topology map model;

[0014] First, the AGVs whose remaining load weight is less than the weight of the goods in the storage area topology map model are eliminated, and then the distance between the driving position of each AGV in the storage area topology map model and the order position is obtained, and the AGV with the shortest distance to the order position is selected, and the AGV is marked as the first AGV;

[0015] Conduct scenario simulation tests on each AGV in advance, and obtain the estimated power consumption of the first AGV to complete the warehouse order delivery process based on the scenario simulation test results, and determine whether the estimated power consumption is greater than the remaining power of the first AGV. If so, mark the first AGV as a key AGV.

[0016] If it is not greater than, the first AGV cart is eliminated, and then the AGV cart with the shortest distance to the order location is re-screened, and the AGV cart is marked as the first AGV cart. According to the scene simulation test results, the estimated power consumption of the first AGV cart to complete the warehouse order delivery process is obtained, and it is determined whether the estimated power consumption is greater than the remaining power of the first AGV cart, and so on, until the key AGV cart is screened out.

[0017] Furthermore, the process of performing scenario simulation tests on each AGV in advance and obtaining the estimated power consumption of the first AGV to complete the warehouse order delivery process according to the scenario simulation test results includes:

[0018] Pre-set several simulation operation scenarios, perform operation tests on each AGV vehicle in several simulation operation scenarios, obtain the unit power consumption of each AGV vehicle under different load capacities and different warehouse and distribution trajectory complexity levels, obtain the load capacities of the first AGV vehicle according to the remaining load capacities and cargo weight of the first AGV vehicle, obtain the warehouse and distribution trajectory complexity level corresponding to the first AGV vehicle, obtain the unit power consumption of the first AGV vehicle under the load capacities and warehouse and distribution trajectory complexity levels, obtain the delivery distance of the first AGV vehicle according to the distance between the driving position of the first AGV vehicle and the order location and the warehouse and distribution trajectory of the warehouse order, and obtain the estimated power consumption of the first AGV vehicle according to the unit power consumption and delivery distance.

[0019] Furthermore, the process of obtaining the complexity level of the warehouse distribution trajectory corresponding to the first AGV vehicle includes:

[0020] Extract the road width and number of obstacles of all road sections in the storage area topology map model, and extract the traffic flow time series of each road section in the historical collection period. Use the traffic flow time series as the test set and training set to build a traffic flow prediction model based on deep learning.

[0021] Input the training set into the traffic flow prediction model for training until the loss function training is stable, and save the model parameters, test the traffic flow prediction model with the test set until it meets the preset requirements, and output the traffic flow prediction model;

[0022] Output the estimated traffic flow of all road sections in the storage area topology map model within the current collection period according to the traffic flow estimation model;

[0023] The section widths, number of obstacles, estimated traffic flow and number of turns of the first AGV in the warehousing and distribution trajectory corresponding to the first AGV are extracted, and the complexity level of the warehousing and distribution trajectory corresponding to the first AGV is obtained according to the section widths, number of obstacles, estimated traffic flow and number of turns of the several sections.

[0024] Furthermore, the process of obtaining the complexity level of the storage and distribution trajectory corresponding to the first AGV according to the width of the road section, the number of obstacles, the estimated traffic flow and the number of turns of the plurality of road sections includes:

[0025] According to the road width, number of obstacles, estimated traffic flow and number of turns of several road sections corresponding to the first AGV car as evaluation indicators, the indicator weight matrix, the complexity level of the warehousing and distribution trajectory and the membership function of the evaluation indicators are preset, and the membership matrix of the evaluation indicators for the complexity level of the warehousing and distribution trajectory is obtained through fuzzy comprehensive evaluation;

[0026] The membership matrix and the indicator weight matrix evaluation indicators are matrix-fused to obtain a fuzzy comprehensive evaluation matrix. The membership of the evaluation indicators for different warehousing and distribution trajectory complexity levels is obtained according to the fuzzy comprehensive evaluation matrix, and the warehousing and distribution trajectory complexity level with the highest membership corresponding to the evaluation indicators is screened out, and the warehousing and distribution trajectory complexity level is used as the warehousing and distribution trajectory complexity level corresponding to the first AGV vehicle.

[0027] Furthermore, a simulation test is performed on the AGV in advance, and a fault matrix set of the key AGV is obtained according to the simulation test result, and the operation data time series sequence of the key AGV in the warehouse order delivery process is monitored in real time. The process of monitoring the operation status of the key AGV according to the operation data time series sequence and the fault matrix set includes:

[0028] Different types of fault scenarios are pre-set, and different types of fault scenarios are tested for each AGV, and the time series of the running data before different types of faults occur on each AGV is obtained. According to the time series of the running data, the stability coefficient matrix of the running data before different types of faults occur on each AGV is extracted, and the stability coefficient matrix of the running data includes the driving position stability coefficient, the driving speed stability coefficient and the remaining power stability coefficient;

[0029] Constructing a task planning table for a key AGV, wherein the task planning table includes a planned driving speed and an estimated power consumption for each section of a storage and distribution track corresponding to the key AGV;

[0030] When the key AGV car starts to deliver warehouse orders, the operation data of the key AGV car is monitored in real time, and the operation data is compared with the task planning table to obtain the operation data stability coefficient matrix of the key AGV car, and the operation data stability coefficient matrix of the key AGV car before different types of failures occur. A fault matrix set is constructed according to the operation data stability coefficient matrix, and the operation data stability coefficient matrix of the key AGV car is matched with the fault matrix set to obtain the similarity between the operation data stability coefficient matrix of the key AGV car and each operation data stability coefficient matrix in the fault matrix set, and a similarity coefficient threshold is preset. If there is an operation data stability coefficient matrix with a similarity greater than the similarity coefficient threshold in the fault matrix set, the fault type corresponding to the operation data stability coefficient matrix is ​​obtained, and a warning signal of the fault type of the key AGV car is generated. If there is no operation data stability coefficient matrix with a similarity greater than the similarity coefficient threshold in the fault matrix set, a normal driving signal of the key AGV car is generated.

[0031] An intelligent debugging and control system for an AGV car for unmanned warehousing, comprising a monitoring center, wherein the monitoring center is communicatively connected to a data acquisition module, an AGV car extraction module and an AGV car monitoring module;

[0032] The data acquisition module is used to build a remote monitoring platform to collect the operation data of the AGV;

[0033] The AGV trolley extraction module is used to construct a storage area topology map model. When the remote monitoring platform receives a storage order, it extracts the key data of the storage order and the operation data of each AGV trolley. According to the positional relationship between each AGV trolley and the storage order in the storage area topology map model and the matching relationship between the operation data of the AGV trolley and the key data, the key AGV trolley is screened out;

[0034] The AGV trolley monitoring module is used to perform simulation tests on the AGV trolley in advance, obtain the fault matrix set of the key AGV trolley according to the simulation test results, monitor the operating data timing sequence of the key AGV trolley in the warehouse order delivery process in real time, and monitor the operating status of the key AGV trolley according to the operating data timing sequence and the fault matrix set.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] 1. When the complexity of the warehouse and distribution trajectory is high, for example: multiple turns: there are multiple consecutive turns in the path, such as shuttling between shelves inside the warehouse, which requires frequent changes of direction; obstacle crossing: there are obstacles that need to be crossed on the path, such as stacked goods or other AGVs; narrow passages: the path passes through narrow passages or crowded areas, and the AGV needs to operate carefully to avoid collisions; these situations will increase the complexity of autonomous path planning and autonomous dynamic scheduling during the driving process of the AGV, resulting in increased power consumption of the AGV. The present invention combines the delivery distance of the warehouse and distribution trajectory and the complexity level of the warehouse and distribution trajectory to determine the actual power consumption of the AGV. Compared with the traditional method of determining the power consumption value of the AGV based on the delivery distance, the accuracy of the power consumption estimation of the AGV is guaranteed.

[0037] 2. Conduct simulation tests on the AGV trolleys in advance, obtain the fault matrix set of the key AGV trolleys based on the simulation test results, monitor the operating data time series sequence of the key AGV trolleys in the warehouse order delivery process in real time, monitor the operating status of the key AGV trolleys based on the operating data time series sequence and the fault matrix set, and take preventive measures before the AGV trolley failure matures, significantly improving the reliability and stability of the AGV trolley warehouse order delivery process. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a schematic diagram of an intelligent debugging and control method for an AGV vehicle used in unmanned warehousing according to an embodiment of the present application.

[0039] Figure 2 This is a schematic diagram of an intelligent debugging and control system for an AGV vehicle used in unmanned warehousing according to an embodiment of the present application. DETAILED DESCRIPTION

[0040] The following is a clear and complete description of the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0041] like Figure 1 As shown, an intelligent debugging and control method for an AGV car for unmanned warehousing includes the following steps:

[0042] Step s1: Build a remote monitoring platform to collect the operation data of the AGV;

[0043] Step s2: Build a storage area topology map model. When the remote monitoring platform receives a storage order, extract the key data of the storage order and the operation data of each AGV. According to the positional relationship between each AGV and the storage order in the storage area topology map model and the matching relationship between the operation data of the AGV and the key data, select the key AGV.

[0044] Step s3: Perform simulation tests on the AGV trolley in advance, obtain the fault matrix set of the key AGV trolley based on the simulation test results, monitor the operating data time series sequence of the key AGV trolley in the warehouse order delivery process in real time, and monitor the operating status of the key AGV trolley based on the operating data time series sequence and the fault matrix set.

[0045] It should be further explained that, in the specific implementation process, the process of building a remote monitoring platform and collecting the operation data of the AGV vehicle includes:

[0046] A remote monitoring platform is constructed, and the remote monitoring platform is connected to a number of data transmission nodes. The data transmission nodes are deployed on the AGV car. The data transmission nodes are used to collect the operating data of the AGV car and mark the collection time and set the collection cycle. The operating data includes the driving position, driving speed, remaining power and remaining load.

[0047] It should be further explained that, in the specific implementation process, a storage area topology map model is constructed. When the remote monitoring platform receives a storage order, the key data of the storage order and the operation data of each AGV are extracted. According to the positional relationship between each AGV and the storage order in the storage area topology map model and the matching relationship between the operation data of the AGV and the key data, the process of screening out the key AGV includes:

[0048] According to the actual situation, build a topological map model of the storage area;

[0049] Key data of warehouse orders include cargo weight, order location, and warehouse delivery track. At the same time, the driving position, remaining power, and remaining load of each AGV are displayed on the warehouse area topology map model.

[0050] First, the AGVs whose remaining load weight is less than the weight of the goods in the storage area topology map model are eliminated, and then the distance between the driving position of each AGV in the storage area topology map model and the order position is obtained, and the AGV with the shortest distance to the order position is selected, and the AGV is marked as the first AGV;

[0051] Conduct scenario simulation tests on each AGV in advance, and obtain the estimated power consumption of the first AGV to complete the warehouse order delivery process based on the scenario simulation test results, and determine whether the estimated power consumption is greater than the remaining power of the first AGV. If so, mark the first AGV as a key AGV.

[0052] If it is not greater than, the first AGV cart is eliminated, and then the AGV cart with the shortest distance to the order location is re-screened, and the AGV cart is marked as the first AGV cart. According to the scene simulation test results, the estimated power consumption of the first AGV cart to complete the warehouse order delivery process is obtained, and it is determined whether the estimated power consumption is greater than the remaining power of the first AGV cart, and so on, until the key AGV cart is screened out.

[0053] It should be further explained that, in the specific implementation process, the process of performing scenario simulation tests on each AGV in advance and obtaining the estimated power consumption of the first AGV to complete the warehouse order delivery process according to the scenario simulation test results includes:

[0054] A number of simulation operation scenarios are pre-set. The simulation operation scenario setting process specifically includes using a simulator to generate a number of different simulation storage orders. A number of different simulation operation scenarios are generated according to the cargo weight, order location and storage and distribution trajectory of each simulation storage order. A number of simulation operation scenarios are run on each AGV to test the unit power consumption of each AGV under different load weights and different levels of storage and distribution trajectory complexity. The load weight of the first AGV is obtained according to the remaining load weight and cargo weight of the first AGV. The load weight of the first AGV = AGV The maximum load capacity - the remaining load capacity of the AGV trolley + the load capacity of the first AGV trolley, obtain the complexity level of the warehouse and distribution trajectory corresponding to the first AGV trolley, obtain the unit power consumption corresponding to the first AGV trolley under the load capacity and the complexity level of the warehouse and distribution trajectory, obtain the delivery distance of the first AGV trolley according to the distance between the driving position of the first AGV trolley and the order position and the warehouse and distribution trajectory of the warehouse order, and obtain the estimated power consumption of the first AGV trolley according to the unit power consumption and the delivery distance; in the present invention, the unit power consumption represents the power consumption value of the AGV trolley for every meter it travels.

[0055] When the complexity of the warehouse and distribution trajectory is high, for example: multiple turns: there are multiple consecutive turns in the path, such as shuttling between shelves inside the warehouse, which requires frequent changes of direction; obstacle crossing: there are obstacles that need to be crossed on the path, such as stacked goods or other AGVs; narrow passages: the path passes through narrow passages or crowded areas, and the AGV needs to operate carefully to avoid collisions; these situations will increase the complexity of autonomous path planning and autonomous dynamic scheduling during the driving process of the AGV, resulting in increased power consumption of the AGV. Therefore, it is necessary to determine the actual power consumption of the AGV in combination with the delivery distance of the warehouse and distribution trajectory and the complexity level of the warehouse and distribution trajectory. Compared with the traditional method of determining the power consumption value of the AGV based on the delivery distance, the present invention ensures the accuracy of the AGV power consumption estimation.

[0056] It should be further explained that, in the specific implementation process, the process of obtaining the complexity level of the warehouse distribution trajectory corresponding to the first AGV vehicle includes:

[0057] Extract the road width and number of obstacles of all road sections in the storage area topology map model, and extract the traffic flow time series of each road section in the historical collection period. Use the traffic flow time series as the test set and training set to build a traffic flow prediction model based on deep learning.

[0058] Input the training set into the traffic flow prediction model for training until the loss function training is stable, and save the model parameters, test the traffic flow prediction model with the test set until it meets the preset requirements, and output the traffic flow prediction model;

[0059] Output the estimated traffic flow of all road sections in the storage area topology map model within the current collection period according to the traffic flow estimation model;

[0060] Extract the width of several road sections included in the storage and distribution track corresponding to the first AGV, the number of obstacles, the estimated traffic flow, and the number of turns of the first AGV in the storage and distribution track, and obtain the complexity level of the storage and distribution track corresponding to the first AGV according to the width of several road sections, the number of obstacles, the estimated traffic flow, and the number of turns;

[0061] Among them, the number of turns indicates the number of times the direction needs to be changed in the warehouse and distribution trajectory. The more turns there are, the higher the complexity level of the warehouse and distribution trajectory.

[0062] The number and type of obstacles indicate the number of obstacles on the path, such as stacked goods, other AGVs, etc. The more obstacles there are, the higher the complexity level of the warehouse and distribution trajectory.

[0063] The road section width indicates the width of the road section. Narrow passages will increase the difficulty of driving. The smaller the road section width, the higher the complexity level of the warehousing and distribution trajectory.

[0064] Traffic flow refers to the number of other AGVs or personnel passing through the road section per unit time. The greater the traffic flow, the higher the complexity level of the warehousing and distribution trajectory.

[0065] It should be further explained that, in the specific implementation process, the process of obtaining the complexity level of the storage and distribution trajectory corresponding to the first AGV vehicle according to the road width, the number of obstacles, the estimated traffic flow and the number of turns of the several road sections includes:

[0066] According to the road width, number of obstacles, estimated traffic flow and number of turns of several road sections corresponding to the first AGV car as evaluation indicators, the indicator weight matrix, the complexity level of the warehousing and distribution trajectory and the membership function of the evaluation indicators are preset, and the membership matrix of the evaluation indicators for the complexity level of the warehousing and distribution trajectory is obtained through fuzzy comprehensive evaluation;

[0067] The membership matrix and the indicator weight matrix evaluation index are matrix-fused to obtain a fuzzy comprehensive evaluation matrix, and the membership of the evaluation index for different warehouse distribution trajectory complexity levels is obtained according to the fuzzy comprehensive evaluation matrix, and the warehouse distribution trajectory complexity level with the highest membership corresponding to the evaluation index is screened out, and the warehouse distribution trajectory complexity level is used as the warehouse distribution trajectory complexity level corresponding to the first AGV vehicle;

[0068] The fuzzy comprehensive evaluation matrix formula obtained by matrix fusion of the membership matrix and the indicator weight matrix evaluation index is:

[0069] ;

[0070] in, is the fuzzy comprehensive evaluation matrix of the evaluation index, is the indicator weight matrix of the evaluation indicator, is the membership matrix, " represents the multiplication of the weight matrix of the evaluation index and the elements at the corresponding positions of the membership matrix, It is a weighted parameter for controlling the balance between the weight matrix and the membership matrix in the fuzzy comprehensive evaluation matrix of the evaluation index.

[0071] It should be further explained that, in the specific implementation process, the AGV car is simulated and tested in advance, and the fault matrix set of the key AGV car is obtained according to the simulation test results, and the operation data time series sequence of the key AGV car in the warehouse order delivery process is monitored in real time. The process of monitoring the operation status of the key AGV car according to the operation data time series sequence and the fault matrix set includes:

[0072] Different types of fault scenarios are pre-set, such as sensor failure, power system failure, path planning error, etc. of the AGV, and operation tests of different types of fault scenarios are performed on each AGV to obtain the time series sequence of operation data of each AGV before different types of faults occur. According to the time series sequence of operation data, the stability coefficient matrix of operation data of each AGV before different types of faults occur is extracted. The stability coefficient matrix of operation data includes the driving position stability coefficient, the driving speed stability coefficient and the remaining power stability coefficient.

[0073] Constructing a task planning table for a key AGV, wherein the task planning table includes a planned driving speed and an estimated power consumption for each section of a storage and distribution track corresponding to the key AGV;

[0074] When the key AGV car starts to deliver warehouse orders, the operation data of the key AGV car is monitored in real time, and the operation data is compared with the task planning table to obtain the operation data stability coefficient matrix of the key AGV car, and the operation data stability coefficient matrix of the key AGV car before different types of failures occur. A fault matrix set is constructed according to the operation data stability coefficient matrix, and the operation data stability coefficient matrix of the key AGV car is matched with the fault matrix set to obtain the similarity between the operation data stability coefficient matrix of the key AGV car and each operation data stability coefficient matrix in the fault matrix set, and a similarity coefficient threshold is preset. If there is an operation data stability coefficient matrix with a similarity greater than the similarity coefficient threshold in the fault matrix set, the fault type corresponding to the operation data stability coefficient matrix is ​​obtained, and a warning signal of the fault type of the key AGV car is generated. If there is no operation data stability coefficient matrix with a similarity greater than the similarity coefficient threshold in the fault matrix set, a normal driving signal of the key AGV car is generated.

[0075] It should be further explained that, in the specific implementation process, the operation data is compared with the task planning table, and the calculation formula for obtaining the operation data stability coefficient matrix of the key AGV vehicle is:

[0076] ;

[0077] ;

[0078] ;

[0079] in, represents the driving position stability coefficient of the i-th road section, ( , , ) represents the three-dimensional coordinates of the key AGV car at the tth moment on the i-th section in the warehouse area topology map model, ( , , ) represents the three-dimensional coordinates of the key AGV car at the tth moment on the i-th section in the storage area topology map model in the task planning table, represents the driving speed stability coefficient of the i-th road section, represents the driving speed of the key AGV car at the tth moment on the i-th section in the warehouse area topology map model, It represents the driving speed of the key AGV car at the tth moment on the i-th section in the storage area topology map model in the task planning table. represents the remaining power stability coefficient of the i-th section, It represents the remaining power of the key AGV vehicle at the tth moment on the i-th section in the warehouse area topology map model. It represents the remaining power of the key AGV car at the tth moment on the i-th section of the storage area topology map model in the task planning table, and n represents the total number of moments.

[0080] The calculation formula for obtaining the similarity between the operating data stability coefficient matrix of the key AGV vehicle and the stability coefficient matrix of each operating data in the fault matrix set is:

[0081] ;

[0082] in, represents the similarity, and m represents the total number of road segments.

[0083] like Figure 2 As shown, an intelligent commissioning and control system for an AGV car for unmanned warehousing includes a monitoring center, wherein the monitoring center is communicatively connected to a data acquisition module, an AGV car extraction module, and an AGV car monitoring module;

[0084] The data acquisition module is used to build a remote monitoring platform to collect the operation data of the AGV;

[0085] The AGV trolley extraction module is used to construct a storage area topology map model. When the remote monitoring platform receives a storage order, it extracts the key data of the storage order and the operation data of each AGV trolley. According to the positional relationship between each AGV trolley and the storage order in the storage area topology map model and the matching relationship between the operation data of the AGV trolley and the key data, the key AGV trolley is screened out;

[0086] The AGV trolley monitoring module is used to perform simulation tests on the AGV trolley in advance, obtain the fault matrix set of the key AGV trolley according to the simulation test results, monitor the operating data timing sequence of the key AGV trolley in the warehouse order delivery process in real time, and monitor the operating status of the key AGV trolley according to the operating data timing sequence and the fault matrix set.

[0087] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An intelligent debugging and control method for an AGV car used in unmanned warehousing, characterized in that: The following steps are involved: Step s1: Build a remote monitoring platform to collect the operation data of the AGV; Step s2: Build a storage area topology map model. When the remote monitoring platform receives a storage order, extract the key data of the storage order and the operation data of each AGV. According to the positional relationship between each AGV and the storage order in the storage area topology map model and the matching relationship between the operation data of the AGV and the key data, select the key AGV. Step s3: Perform simulation tests on the AGV trolleys in advance, obtain the fault matrix set of the key AGV trolleys according to the simulation test results, monitor the operating data time series sequence of the key AGV trolleys in the warehouse order delivery process in real time, and monitor the operating status of the key AGV trolleys according to the operating data time series sequence and the fault matrix set; The process of building a remote monitoring platform and collecting the operating data of the AGV includes: Construct a remote monitoring platform, which is connected to a number of data transmission nodes, which are deployed on the AGV. The data transmission nodes are used to collect the operation data of the AGV and mark the collection time, and set the collection cycle. The operation data includes the driving position, driving speed, remaining power and remaining load; Construct a topological map model of the storage area. When the remote monitoring platform receives a storage order, it extracts the key data of the storage order and the operation data of each AGV. According to the positional relationship between each AGV and the storage order in the topological map model of the storage area and the matching relationship between the operation data of the AGV and the key data, the process of screening out the key AGV includes: According to the actual situation, build a topological map model of the storage area; The key data of the warehouse order includes the weight of the goods, the order location and the warehouse distribution track. At the same time, the driving position, remaining power and remaining load of each AGV are displayed in the warehouse area topology map model; First, the AGVs whose remaining load weight is less than the weight of the goods in the storage area topology map model are eliminated, and then the distance between the driving position of each AGV in the storage area topology map model and the order position is obtained, and the AGV with the shortest distance to the order position is selected, and the AGV is marked as the first AGV; Conduct scenario simulation tests on each AGV in advance, and obtain the estimated power consumption of the first AGV to complete the warehouse order delivery process based on the scenario simulation test results, and determine whether the estimated power consumption is greater than the remaining power of the first AGV. If so, mark the first AGV as a key AGV. If not, the first AGV is eliminated, and then the AGV with the shortest distance to the order location is re-screened, and the AGV is marked as the first AGV. The estimated power consumption of the first AGV to complete the warehouse order delivery process is obtained according to the scenario simulation test results, and it is determined whether the estimated power consumption is greater than the remaining power of the first AGV, and so on, until the key AGV is screened out; The process of performing scenario simulation tests on each AGV in advance and obtaining the estimated power consumption of the first AGV to complete the warehouse order delivery process according to the scenario simulation test results includes: Pre-set several simulation operation scenarios, perform operation tests of several simulation operation scenarios on each AGV, obtain the unit power consumption of each AGV under different load weights and different storage and distribution trajectory complexity levels, obtain the load weight of the first AGV according to the remaining load weight and cargo weight of the first AGV, obtain the storage and distribution trajectory complexity level corresponding to the first AGV, obtain the unit power consumption of the first AGV under the load weight and storage and distribution trajectory complexity level, obtain the delivery distance of the first AGV according to the distance between the driving position of the first AGV and the order position and the storage and distribution trajectory of the storage order, and obtain the estimated power consumption of the first AGV according to the unit power consumption and delivery distance; The process of obtaining the complexity level of the warehouse distribution trajectory corresponding to the first AGV vehicle includes: Extract the road width and number of obstacles of all road sections in the storage area topology map model, and extract the traffic flow time series of each road section in the historical collection period. Use the traffic flow time series as the test set and training set to build a traffic flow prediction model based on deep learning. Input the training set into the traffic flow prediction model for training until the loss function training is stable, and save the model parameters, test the traffic flow prediction model with the test set until it meets the preset requirements, and output the traffic flow prediction model; Output the estimated traffic flow of all road sections in the storage area topology map model within the current collection period according to the traffic flow estimation model; Extract the width of several road sections included in the storage and distribution track corresponding to the first AGV, the number of obstacles, the estimated traffic flow, and the number of turns of the first AGV in the storage and distribution track, and obtain the complexity level of the storage and distribution track corresponding to the first AGV according to the width of several road sections, the number of obstacles, the estimated traffic flow, and the number of turns; The process of obtaining the complexity level of the storage and distribution trajectory corresponding to the first AGV vehicle according to the width of the road section, the number of obstacles, the estimated traffic flow and the number of turns of the plurality of road sections includes: According to the road width, number of obstacles, estimated traffic flow and number of turns of several road sections corresponding to the first AGV car as evaluation indicators, the indicator weight matrix, the complexity level of the warehousing and distribution trajectory and the membership function of the evaluation indicators are preset, and the membership matrix of the evaluation indicators for the complexity level of the warehousing and distribution trajectory is obtained through fuzzy comprehensive evaluation; The membership matrix and the indicator weight matrix evaluation index are matrix-fused to obtain a fuzzy comprehensive evaluation matrix, and the membership of the evaluation index for different warehouse distribution trajectory complexity levels is obtained according to the fuzzy comprehensive evaluation matrix, and the warehouse distribution trajectory complexity level with the highest membership corresponding to the evaluation index is screened out, and the warehouse distribution trajectory complexity level is used as the warehouse distribution trajectory complexity level corresponding to the first AGV vehicle; Conduct simulation tests on the AGV in advance, obtain the fault matrix set of the key AGV according to the simulation test results, monitor the operation data time series of the key AGV in the warehouse order delivery process in real time, and monitor the operation status of the key AGV according to the operation data time series and the fault matrix set. The process includes: Different types of fault scenarios are pre-set, and different types of fault scenarios are tested for each AGV, and the time series of the running data before different types of faults occur on each AGV is obtained. According to the time series of the running data, the stability coefficient matrix of the running data before different types of faults occur on each AGV is extracted, and the stability coefficient matrix of the running data includes the driving position stability coefficient, the driving speed stability coefficient and the remaining power stability coefficient; Constructing a task planning table for a key AGV, wherein the task planning table includes a planned driving speed and an estimated power consumption for each section of a storage and distribution track corresponding to the key AGV; When the key AGV car starts to deliver warehouse orders, the operation data of the key AGV car is monitored in real time, and the operation data is compared with the task planning table to obtain the operation data stability coefficient matrix of the key AGV car, and the operation data stability coefficient matrix of the key AGV car before different types of failures occur, and a fault matrix set is constructed according to the operation data stability coefficient matrix, and the operation data stability coefficient matrix of the key AGV car is matched with the fault matrix set to obtain the similarity between the operation data stability coefficient matrix of the key AGV car and each operation data stability coefficient matrix in the fault matrix set, and a similarity coefficient threshold is preset. If there is an operation data stability coefficient matrix with a similarity greater than the similarity coefficient threshold in the fault matrix set, the fault type corresponding to the operation data stability coefficient matrix is ​​obtained, and a warning signal of the fault type of the key AGV car is generated. If there is no operation data stability coefficient matrix with a similarity greater than the similarity coefficient threshold in the fault matrix set, a normal driving signal of the key AGV car is generated; An intelligent debugging and control method for an AGV car for unmanned warehousing, characterized in that it includes a monitoring center, wherein the monitoring center is communicatively connected to a data acquisition module, an AGV car extraction module and an AGV car monitoring module; The data acquisition module is used to build a remote monitoring platform to collect the operation data of the AGV; The AGV trolley extraction module is used to construct a storage area topology map model. When the remote monitoring platform receives a storage order, it extracts the key data of the storage order and the operation data of each AGV trolley. According to the positional relationship between each AGV trolley and the storage order in the storage area topology map model and the matching relationship between the operation data of the AGV trolley and the key data, the key AGV trolley is screened out; The AGV trolley monitoring module is used to perform simulation tests on the AGV trolley in advance, obtain the fault matrix set of the key AGV trolley according to the simulation test results, monitor the operating data time series sequence of the key AGV trolley in the warehouse order delivery process in real time, and monitor the operating status of the key AGV trolley according to the operating data time series sequence and the fault matrix set; The unit power consumption refers to the power consumption value of the AGV car for every meter it travels.