Intelligent test adjustment method, electronic device, system, and storage medium

Abstract

The application provides a kind of intelligent test adjustment method, electronic equipment, system and storage medium, for the test of automatic guided vehicle, the method includes setting test item and test environment, according to test item setting chain test instruction;Quantitative processing is carried out to the test content of test instruction, and the expected behavior vector is formed;The actual behavior vector of automatic guided vehicle is obtained by executing test instruction;The matching degree between expected behavior vector and actual behavior vector is calculated;When the matching degree is lower than the preset value, deviation is corrected based on the preset knowledge base.The technical scheme of the application realizes performance test and also realizes error correction function.The scheme realizes the automatic performance test of AGV, and intelligent deviation correction is carried out, which greatly improves the operation accuracy of automatic guided vehicle, improves the test effect and stability.

Description

Technical Field

[0001] This application relates to the field of intelligent automation equipment technology, and in particular to intelligent testing and adjustment methods, electronic devices, systems and storage media. Background Technology

[0002] Automated Guided Vehicles (AGVs), also known as automated guided vehicles or automated guided transport vehicles, are industrial vehicles that automatically or manually load goods and travel along a pre-set route to a designated location, where goods are then automatically or manually loaded and unloaded. The continuous development of computer hardware technology, parallel and distributed processing technology, automatic control technology, sensor technology, and software development environments has provided the necessary technological foundation for AGV research and application. The development of artificial intelligence technologies, such as understanding and search, task and path planning, and fuzzy and neural network control technologies, is driving AGVs towards greater intelligence and autonomy. AGV research and development integrates artificial intelligence, information processing, and image processing, involving multiple disciplines such as computer science, automatic control, information communication, mechanical design, and electronic technology, making it one of the hot topics in logistics automation research.

[0003] Despite years of research on AGVs, several key technologies still require improvement and breakthroughs to further enhance AGV performance and reduce manufacturing and operating costs. In this process, testing the accuracy, stability, and other performance aspects of the AGVs to ensure their high-quality and efficient operation is particularly important. Current testing methods often rely on manual settings, with each test performed individually. When problems are discovered, there is no automated troubleshooting mechanism; the results often depend on the experience and judgment of the testers, leading to low efficiency, poor testing results, and instability. Summary of the Invention

[0004] To help improve the problems of low efficiency, poor test results and poor stability, this application provides an intelligent test adjustment method, electronic device, system and storage medium.

[0005] Firstly, this application provides an intelligent test adjustment method for use in the testing of automated guided vehicles, the method comprising:

[0006] Set up test projects and test environments, and set up chained test instructions according to the test projects;

[0007] The test instructions for the test content are quantified to form the expected behavior vector;

[0008] Execute the test command to obtain the actual behavior vector of the automated guided vehicle;

[0009] Calculate the matching degree between the expected behavior vector and the actual behavior vector;

[0010] When the matching degree is lower than the preset value, correction is performed based on the preset knowledge base.

[0011] Secondly, this application provides an electronic device that adopts the following technical solution:

[0012] An electronic device comprising:

[0013] At least one processor;

[0014] Memory;

[0015] At least one application, wherein the at least one application is stored in memory and configured to be executed by at least one processor, the at least one application being configured to: execute the intelligent test tuning method as described above.

[0016] Thirdly, this application provides an intelligent testing system, which includes: an automated guided vehicle testing site, testing hardware and electronic equipment, wherein the electronic equipment is used to execute the intelligent testing adjustment method described above.

[0017] Fourthly, this application provides a computer-readable storage medium, which adopts the following technical solution:

[0018] A computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to perform any of the intelligent test adjustment methods provided in the first aspect.

[0019] In summary, the technical solution of this application is used to perform performance testing of automated guided vehicles (AGVs). It pre-quantifies the test indicators of the test content to form expected behavior vectors, then obtains the actual behavior vectors of the vehicle body, and calculates the matching degree between the expected and actual behavior vectors, thus realizing performance testing. Furthermore, it can adjust the test data based on a knowledge base to maximize the matching degree and achieve error correction. This solution realizes automated performance testing of AGVs and performs intelligent correction, greatly improving the operational accuracy of automated guided vehicles and enhancing test results and stability. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating an intelligent test adjustment method provided in an embodiment of this application;

[0021] Figure 2 This is a schematic diagram of a two-layer network architecture for an intelligent test adjustment method provided in an embodiment of this application;

[0022] Figure 3This is a partial flowchart illustrating an intelligent test adjustment method provided in another embodiment of this application;

[0023] Figure 4 This is a schematic diagram of the hardware performance factor processing flow of an intelligent test adjustment method provided in another embodiment of this application;

[0024] Figure 5 This is a schematic diagram of the environmental factor processing flow of an intelligent test adjustment method provided in another embodiment of this application;

[0025] Figure 6 This is a schematic diagram of the processing flow of the software configuration factor part of an intelligent test adjustment method provided in another embodiment of this application;

[0026] Figure 7 This is a schematic diagram of the module principle of an electronic device provided in an embodiment of this application. Detailed Implementation

[0027] To make the purpose, technical solution, and advantages of this application clearer, the following description is provided in conjunction with the appendix. Figure 1-7 The present application will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the application.

[0028] This application provides an intelligent test adjustment method for use in the testing of automated guided vehicles (AGVs) within the system.

[0029] This embodiment uses an automated guided vehicle (AGV) testing system as an application example to specifically illustrate the above method. In actual implementation, the above method can also be used for testing other automated equipment, and this embodiment does not limit it to this.

[0030] Specific reference Figure 1 As shown, this invention proposes an intelligent test adjustment method, the method comprising:

[0031] S10: Set up test items and test environment, and set up chained test instructions according to the test items;

[0032] S20: Quantify the test instructions of the test content to form the expected behavior vector;

[0033] S30: Execute test instructions to obtain the actual behavior vector of the automated guided vehicle;

[0034] S40: Calculate the matching degree between the expected behavior vector and the actual behavior vector;

[0035] S50: When the matching degree is lower than the preset value, correction is performed based on the preset knowledge base.

[0036] This application's technical solution is used to perform performance testing on automated guided vehicles (AGVs). Performance testing is achieved by calculating the matching degree between the expected behavior vector and the actual behavior vector. Furthermore, it can adjust the test data based on a knowledge base to maximize the matching degree, thus implementing error correction. This solution realizes automated performance testing of AGVs and performs intelligent correction, significantly improving the operational accuracy of the AGVs and enhancing the testing results and stability.

[0037] In this embodiment of the invention, the method is implemented through a two-layer comparison architecture. The first layer of the two-layer comparison architecture compares the matching degree between the actual behavior vector and the expected behavior vector of the test item through a two-layer network. When the matching degree is lower than a preset value, the second layer of the two-layer comparison architecture compares and searches for problem solutions through a knowledge base composed of three two-layer networks, and repairs the anomaly.

[0038] Understandably, this solution employs a two-layer network, two-layer comparison architecture. The first layer compares the matching degree between the current actual behavior vector and the expected behavior vector of the test item, implemented through a two-layer network. When the matching degree between the current actual behavior vector and the expected behavior vector is low, after locating the behavior vector element that triggers the low matching degree, a second layer comparison is performed through a knowledge base composed of three two-layer networks, and a problem solution is searched for and matched to correct the cause of the anomaly.

[0039] See attached document Figure 2 To be continued Figure 6 As shown, based on this, the first and second layers of the two-layer contrast architecture are both designed in vector format, and each element is an element of a vector structure.

[0040] In this dual-layer comparison architecture, each element x_i in the first layer has a connection weight w_ij with each element y_j in the second layer;

[0041] Let the initial value of the connection weight w_ij be 1, and let the connection weight w_ij be updated by the following formula: w_ij new =β*(I ⋀w_ij old )+(1-β)* w_ij old ;

[0042] Among them, w_ij old The connection weights before the update, w_ij new Here, I represents the updated connection weights, β represents the actual behavior vector, and β is a parameter that takes values ​​from 0 to 1.

[0043] The following formula is used to calculate whether the ratio of the similarity between the current actual behavior vector I and the connection weight w_ij relative to the current actual behavior vector is greater than or equal to a threshold. :| ;in For parameters whose values ​​are in the range of 0 to 1:

[0044] If greater than or equal to the threshold If the value is less than the threshold, then return the updated connection weight w_ij; Then, the mismatched behavior structure indicator element will be located.

[0045] Understandably, the matching degree between the current actual behavior vector and the expected behavior vector of the test item is calculated first. If the matching degree is lower than the threshold, the matching degree is considered lower. Condition, | Then the network will make corrections based on the knowledge base.

[0046] See attached document Figure 2 As shown, the first layer of the architecture is designed in vector format, where each element represents the current actual behavior vector structure element, with I{I_1, I_1···I_n} representing a current actual behavior vector;

[0047] The first layer of the architecture is designed in vector format, with each element encoding a desired behavior vector, represented by T={T_1(I)···T_M(I)};

[0048] Each element in the first layer is connected to each element in the second layer by w_ij. The current actual behavior vector I is encoded by the connection weight vector w_j to form the corresponding expected behavior vector T_j(I).

[0049] Before encoding, each current actual behavior vector is compared with an existing expected behavior vector. If the match is high, the expected behavior vector corresponding to the existing current actual behavior vector is called. ;

[0050] If the matching degree is low, then a new expected behavior vector corresponding to the current actual behavior vector is created:

[0051] .

[0052] Understandably, since the initial input data for the network is the expected behavior vector for each test item, the expected behavior vector for each test item is encoded in the network by updating the connection weights w_j. When the current actual behavior vector I matches the existing expected behavior vector with sufficiently high accuracy, ;in , representing the expected behavior vector corresponding to the current actual behavior vector.

[0053] If the current actual behavior vector I has a low matching degree with all expected behavior vectors, then a new element is created. .

[0054] In this embodiment of the invention, the knowledge base includes three two-layer networks adapted to hardware factors, software configuration factors, and environmental factors, respectively; the behavioral structure index elements include hardware factors, software configuration factors, and environmental factors. Of course, the knowledge base may also include other factors and corresponding networks, which are not limited to the description in this embodiment.

[0055] Building upon this, the steps executed by the second layer of the two-layer contrastive architecture's network further include:

[0056] Update the connection weights;

[0057] The similarity between the actual behavior vector and the connection weights is calculated for each cycle;

[0058] If the similarity meets the standard, select the existing element with the highest similarity; if the similarity does not meet the standard, create a new element.

[0059] Perform automatic software repair or notify manual repair.

[0060] It should be noted that the aforementioned manual repair refers to the pairing of output problem solutions between two two-layer networks in the knowledge base, as shown in the appendix. Figures 3 to 6 When the knowledge base's two-layer network for "Hardware Performance" is invoked, the output solution requires manual inspection of the hardware, parameter modification, and retesting to obtain a new current actual behavior vector. Similarly, when the knowledge base's two-layer network for "Environmental Factors" is invoked, the output solution requires manual inspection of environmental factors and retesting to obtain a new current actual behavior vector.

[0061] In an embodiment of the present invention:

[0062] The data in the knowledge base network is a vector of problem solution results, with the following structure:

[0063] Hardware performance, ={Indicator element, hardware device, hardware parameter};

[0064] Environmental factors ={Indicator elements, environmental factors, and change measures};

[0065] and software configuration, ={Indicator elements, software algorithm, software parameters};

[0066] Among them, the indicator element is the problem, and the hardware equipment, hardware parameters, environmental factors, change measures, software algorithm and software parameters are the solutions to the problem;

[0067] When the match between the actual behavior vector and the expected behavior vector is too low and mismatch element localization is triggered, the knowledge base network traverses the following input data until the match meets the requirements:

[0068] Hardware performance, ={Indicator element, non-measured device, non-measured parameter};

[0069] Environmental factors ={Indicator elements, non-measured environmental factors, non-measured parameters};

[0070] Software configuration, ={Indicator element, non-measured algorithm, non-measured parameter};

[0071] The knowledge base network output is a problem solution structure vector.

[0072] Understandably, the knowledge base of this solution consists of three two-layer networks, representing three main reasons why the expected behavior vector of each test item does not match the current actual behavior vector: [hardware performance], [software configuration], and [environmental factors]. The network computation process is illustrated with three case studies; please see the appendix. Figures 3 to 6 .

[0073] Example 1, Speed ​​Mismatch [Hardware Performance]: Assuming the expected behavior vector is derived from the test item parsing...

[0074] {Action type: Forward (or moving)}

[0075] Target value: The specified speed (e.g., 1.0 m / s)

[0076] Unit: m / s (meters per second)

[0077] Tolerance: ±0.1m / s

[0078] Indicator elements: Actual speed vs. Expected speed

[0079] When the matching degree between the expected behavior vector and the current actual behavior vector is lower than the threshold The mismatched indicator element is identified as the actual speed not meeting the expected speed. Therefore, the knowledge base's two-layer network for hardware performance is invoked to perform the second layer comparison of the solutions.

[0080] When the expected behavior vector does not match the current actual behavior vector for the first time, the input data for the knowledge base's two-layer network (hardware performance) is...

[0081] I^hdw=[Speed ​​mismatch, any device, any parameter],

[0082] If the second-level comparison condition is met, the maximum value T^hdw_J is automatically selected; if the second-level comparison condition is not met, a new element T_j(I^hdw) is created.

[0083] When the second expected behavior vector does not match the current actual behavior vector, or the mismatch persists, the input data for the knowledge base's two-layer network with [hardware performance] is...

[0084] I^hdw=[Speed ​​mismatch, non-selected hardware, any parameter],

[0085] Based on the second-level comparison conditions, existing problem solutions are invoked for pairing or new elements are created.

[0086] The knowledge base's example of matching solutions for the two-layer network output problem related to "hardware performance" can be...

[0087] T_J(I^hdw)=[Speed ​​mismatch, check motor, check motor power],

[0088] T_J(I^hdw)=[Speed ​​mismatch, check battery, check battery level],

[0089] Alternatively, a solution can be found for other hardware-related issues.

[0090] The solution output for the [hardware performance] problem requires manual intervention, that is, notifying the testers to check the specified parameters of the specified hardware.

[0091] After checking the hardware and modifying the parameters, the test is repeated to obtain a new current actual behavior vector. The network's two-layer comparison architecture is used to perform the first comparison with the expected behavior vector of the test item. If the matching degree is still low, the three two-layer networks of the knowledge base are used for a second comparison and search to determine the matching of the problem solution that matches the expected behavior vector.

[0092] In Example 2, the positioning accuracy error [software configuration] is assumed to be derived from the expected behavior vector parsed from the test item.

[0093] {Action type: Positioning (using sensors for positioning)}

[0094] Target value: Precisely located at a specific coordinate point (e.g., point A).

[0095] Unit: meter (m)

[0096] Tolerance: ±5mm

[0097] Indicator elements: Actual coordinates vs. target coordinates}

[0098] When the matching degree between the expected behavior vector and the current actual behavior vector is lower than the threshold The mismatched indicator elements are those whose actual coordinates do not match the target coordinates. Therefore, the knowledge base's two-layer network of "Software Configuration" is invoked to perform the second layer comparison of the solutions.

[0099] When the expected behavior vector does not match the current actual behavior vector for the first time, the input data for the two-layer network of the knowledge base's [software configuration] is:

[0100] I^sfw = [Positioning accuracy error, any algorithm, any parameters]

[0101] If the second-level comparison condition is met, the maximum value T^sfw_J is automatically selected; if the second-level comparison condition is not met, a new element T_j(I^sfw) is created.

[0102] When the second expected behavior vector does not match the current actual behavior vector, or the mismatch persists, the input data for the knowledge base's two-layer network in the [software configuration] is:

[0103] I^sfw=[Positioning accuracy error, non-selected algorithm, any parameter],

[0104] Based on the second-level comparison conditions, existing problem solutions are invoked for pairing or new elements are created.

[0105] The knowledge base's "Software Configuration" example for matching solutions to a two-layer network output problem can be...

[0106] T_J(I^sfw) = [Positioning accuracy error, check SLAM algorithm, check configuration parameters]

[0107] T_J(I^sfw) = [Velocity mismatch, check sensor fusion algorithm, check fusion parameters],

[0108] Alternatively, a solution can be found for other software-related issues.

[0109] The solution output for the [Software Configuration] problem does not require manual intervention and can be automatically repaired by modifying parameters.

[0110] After checking the software and modifying the configuration parameters, the test is repeated to obtain the new current actual behavior vector. The first comparison is then performed again using the two-layer comparison architecture of the network with the expected behavior vector of the test item. If the matching degree remains low, a second comparison is performed using the three two-layer networks of the knowledge base. The search determines the matching of the problem solution that matches the expected behavior vector.

[0111] In Example 3, ambient light affects sensor accuracy [environmental factors]: Assuming the expected behavior vector is derived from the test item analysis:

[0112] {Action Type: Obstacle Avoidance}

[0113] Target value: Avoid obstacles

[0114] Unit: meter (m)

[0115] Tolerance: ±10mm

[0116] Indicator elements: Actual obstacle avoidance distance vs. Expected obstacle avoidance distance.

[0117] When the matching degree between the expected behavior vector and the current actual behavior vector is lower than the threshold The mismatched indicator element is that the actual obstacle avoidance distance does not meet the expected obstacle avoidance distance. Therefore, the second layer of the solution comparison is performed by calling the two-layer network of the knowledge base, namely the "Environmental Factors".

[0118] When the expected behavior vector does not match the current actual behavior vector, the input data for the knowledge base's two-layer network of "Environmental Factors" is:

[0119] I^env = [Ambient light affects sensor accuracy; any environmental factor, any parameter].

[0120] If the second-level comparison condition is met, the maximum value T^env_J is automatically selected; if the second-level comparison condition is not met, a new element T_j(I^env) is created.

[0121] When the second expected behavior vector does not match the current actual behavior vector, or the mismatch persists, the input data for the knowledge base's two-layer network of "Environmental Factors" is:

[0122] I^env = [Ambient light affects sensor accuracy; any parameter not selected for environmental factors].

[0123] Based on the second-level comparison conditions, existing problem solutions are invoked for pairing or new elements are created.

[0124] The example of matching solutions to the two-layer network output problem in the knowledge base for "Environmental Factors" can be...

[0125] T_J(I^env)=[Ambient light affects sensor accuracy; check ambient light; check its impact on lidar].

[0126] T_J(I^env) = [Ambient light affects sensor accuracy; check ambient light; check its impact on the camera].

[0127] Matching solutions to other environmental factors or related issues.

[0128] The solution output for the [Environmental Factors] issue requires manual intervention, that is, notifying testers to check the specified environmental factors and their impact on the corresponding hardware.

[0129] After checking environmental factors, the test is retested to obtain a new current actual behavior vector. The two-layer comparison architecture of the network is used to perform the first comparison with the expected behavior vector of the test item. If the matching degree is still low, the three two-layer networks of the knowledge base are used to perform a second comparison, search, and determine the matching of the problem solution that meets the expected behavior vector.

[0130] It should be noted that since the initial input data for the network is the expected behavior vector for each test item, the expected behavior vector for each test item is encoded in the network by updating the connection weights w_j. When the current actual behavior vector I matches the existing expected behavior vector sufficiently well...

[0131]

[0132] in

[0133] This represents the expected behavior vector corresponding to the current actual behavior vector.

[0134] If the current actual behavior vector I has a low matching degree with all expected behavior vectors, then a new element is created:

[0135] .

[0136] Understandably, if the first-level comparison reveals a low match between the current actual behavior vector and the expected behavior vector, the reason is identified as the indicator elements of the behavior structure not meeting the expected values. A high match indicates that the indicator elements meet the expected values. The second-level comparison then searches for and determines the matching solutions that match the expected behavior vector. Therefore, the first-level comparison in the solution process is used to locate behavioral structure factors, the second-level comparison is used to determine the matching solutions based on these behavioral structure factors, and the match degree comparison is performed again in the above manner.

[0137] This application also provides an electronic device, such as... Figure 7 As shown, Figure 7 The illustrated electronic device 700 includes a processor 701 and a memory 703. The processor 701 and the memory 703 are connected, for example, via a bus 702. Optionally, the electronic device 700 may also include a transceiver 704. It should be noted that in practical applications, the transceiver 704 is not limited to one type, and the structure of this electronic device 700 does not constitute a limitation on the embodiments of this application.

[0138] Processor 701 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), or other programmable logic devices, transistor logic devices, hardware components, or any other combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 701 may also be a combination that implements computing functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0139] Bus 702 may include a pathway for transmitting information between the aforementioned components. Bus 702 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Bus 702 can be divided into address bus, data bus, etc. For ease of representation, Figure 7 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0140] The memory 703 may be ROM (Read Only Memory) or other types of static storage devices capable of storing static information and instructions, RAM (Random Access Memory) or other types of dynamic storage devices capable of storing information and instructions, or EEPROM (Electrically Erasable Programmable Read Only Memory), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto.

[0141] The memory 703 is used to store application code that executes the solution of this application, and its execution is controlled by the processor 701. The processor 701 is used to execute the application code stored in the memory 703 to implement the content shown in the foregoing method embodiments.

[0142] Electronic devices include, but are not limited to: mobile terminals such as mobile phones, laptops, PDAs (personal digital assistants), and PADs (tablet computers), as well as fixed terminals such as digital TVs and desktop computers. Servers can also be included. Figure 7 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0143] This application also provides an intelligent testing system, which includes: an automated guided vehicle testing site, testing hardware and electronic equipment, wherein the electronic equipment is used to execute the intelligent testing adjustment method described above.

[0144] Its test items may include the following:

[0145] driving in a straight line Proceed along a straight path Speed, path deviation, running time stop Stop at the target point Position error, stopping error, time Turn Rotate at a specified angle Angular deviation, turning radius, angular velocity Obstacle Avoidance Adjust path when encountering obstacles Obstacle avoidance trigger time and trajectory after obstacle avoidance Acceleration / Deceleration speed change Acceleration, velocity curve Charge Move to a charging station and connect to a charger. Charging position error, contact success rate

[0146] In calculating the current actual behavior vector, features need to be extracted from the collected sensor data. These features are extracted based on the expected behavior measurement indicators to evaluate whether the AGV has achieved the expected performance.

[0147] First, the feature recognition of the current behavior is mainly based on the following sensor data:

[0148] Odometer / UWB Position (X, Y, θ) m, rad Calculate path and deviation Wheel speed gauge speed m / s Calculate running status IMU acceleration m / s² Identify acceleration and deceleration actions IMU angular velocity rad / s Identify steering and rotation movements Camera Visual data Pixels Identify tracks and obstacles LiDAR Distance, obstacles m Calculate obstacle avoidance conditions

[0149] The core features extracted include

[0150] Location characteristics: current position (X,Y,θ), motion trajectory (path deviation);

[0151] Velocity characteristics: linear velocity v (m / s), angular velocity ω (rad / s);

[0152] Acceleration characteristics: linear acceleration a (m / s²), angular acceleration α (rad / s²);

[0153] Time characteristics: Task execution time T(s), timestamps of key actions;

[0154] Error characteristics: deviations ΔX, ΔY (m), angular error Δθ (rad);

[0155] Environmental characteristics (optional): Road smoothness, obstacle distribution;

[0156] The following features are then converted into quantifiable metrics to calculate the current actual behavior vector:

[0157] Displacement deviation Deviation between the actual path and the expected path of the AGV ΔX = X_actual - X_expected Angle deviation Rotation angle error Δθ = θ_actual - θ_expected Final docking error Error between the AGV target point and the actual stopping point √(ΔX² + ΔY²) velocity curve velocity as a function of time v(t) = dx / dt Acceleration change acceleration curve over time a(t) = dv / dt runtime Task execution time T = t_end - t_start Path tracking error Error between the running trajectory and the reference trajectory ∑ √((X_actual - X_ref)² + (Y_actual - Y_ref)²) / N

[0158] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed in a computer, causes the computer to perform the intelligent test adjustment method provided in the above embodiments.

[0159] It should be understood that although the steps in the flowcharts in the accompanying drawings are shown sequentially as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise expressly stated herein, there is no strict order in which these steps are performed, and they may be performed in other orders.

[0160] The above are only some embodiments of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. An intelligent test adjustment method, characterized in that, In testing of automated guided vehicles, the method includes: Set up test projects and test environments, and set up chained test instructions according to the test projects; The test instructions for the test content are quantified to form the expected behavior vector; Execute the test command to obtain the actual behavior vector of the automated guided vehicle; Calculate the matching degree between the expected behavior vector and the actual behavior vector; When the matching degree is lower than the preset value, correction is performed based on the preset knowledge base; This also includes a two-layer comparison architecture. The first layer of the two-layer comparison architecture compares the matching degree between the actual behavior vector and the expected behavior vector of the test item through a two-layer network. When the matching degree is lower than a preset value, the second layer of the two-layer comparison architecture compares and searches for problem solutions through a knowledge base composed of three two-layer networks, and repairs the anomaly. In the two-layer contrastive architecture, each element x_i in the first layer has a connection weight w_ij with each element y_j in the second layer; Let the initial value of the connection weight w_ij be 1, and let the connection weight w_ij be updated by the following formula: w_ij new =β *(I⋀w_ij old )+(1-β)* w_ij old ; Among them, w_ij old The connection weights before the update, w_ij new Here, I represents the updated connection weights, β represents the actual behavior vector, and β is a parameter that takes values ​​from 0 to 1. The following formula is used to calculate whether the ratio of the similarity between the current actual behavior vector I and the connection weight w_ij relative to the current actual behavior vector is greater than or equal to a threshold. :| ;in For parameters whose values ​​are in the range of 0 to 1: If greater than or equal to the threshold If the value is less than the threshold, then return the updated connection weight w_ij; Then locate the mismatched behavior structure indicator element; The knowledge base includes three two-layer networks that are adapted to hardware factors, software configuration factors, and environmental factors, respectively; the behavioral structure index elements include hardware factors, software configuration factors, and environmental factors. The steps performed by the second layer of the dual-layer network in the dual-layer contrast architecture include: Update the connection weights; The similarity between the actual behavior vector and the connection weights is calculated for each cycle; If the similarity meets the standard, select the existing element with the highest similarity; if the similarity does not meet the standard, create a new element. Perform automatic software repair or notify manual repair.

2. The intelligent test adjustment method according to claim 1, characterized in that, The first and second layers of the dual-layer comparison architecture are both designed in vector format, and each element is an element of a vector structure.

3. The intelligent test adjustment method according to claim 1, characterized in that, in: The data in the knowledge base network is a vector of problem solution results, with the following structure: Hardware performance, ={Indicator element, hardware device, hardware parameter}; Environmental factors ={Indicator elements, environmental factors, and change measures}; and software configuration, ={Indicator elements, software algorithm, software parameters}; Among them, the indicator element is the problem, and the hardware equipment, hardware parameters, environmental factors, change measures, software algorithm and software parameters are the solutions to the problem; When the match between the actual behavior vector and the expected behavior vector is too low and mismatch element localization is triggered, the knowledge base network traverses the following input data until the match meets the requirements: Hardware performance, ={Indicator element, non-measured device, non-measured parameter}; Environmental factors ={Indicator elements, non-measured environmental factors, non-measured parameters}; Software configuration, ={Indicator element, non-measured algorithm, non-measured parameter}; The knowledge base network output is a problem solution structure vector.

4. An electronic device, characterized in that, The electronic device includes: At least one processor; Memory; At least one application, wherein the at least one application is stored in memory and configured to be executed by at least one processor, said at least one application being configured to: perform the intelligent test adjustment method according to any one of claims 1 to 3.

5. An intelligent testing system, characterized in that, The intelligent testing system includes: an automated guided vehicle testing area, testing hardware and electronic equipment, wherein the electronic equipment is used to execute the intelligent testing adjustment method as described in any one of claims 1 to 3.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed in the computer, the computer is instructed to perform the intelligent test adjustment method according to any one of claims 1 to 3.

Images