Military crane chassis fault reproduction and practical training system based on working condition simulation

Abstract

The invention discloses a military crane chassis fault reproduction and practical training system based on working condition simulation, and relates to the technical field of military equipment practical training, and the system comprises an entity chassis building unit which selects and uses entity parts of subsystems of a military automobile crane to be arranged and installed according to a real automobile, retains a key relation, presets an interface, defines a safety threshold value and the like; the dual-mode fault generation unit constructs an initial library, and an intelligent agent is embedded to generate a fault; the multi-dimensional data acquisition unit responds to the trigger signal to acquire multiple types of data; the fault diagnosis practical operation unit is used for trainees to operate; the training evaluation and fault iteration unit generates a report and feeds back optimization suggestions; according to the invention, a dual-mode fault generation unit is constructed, an initial fault library and an intelligent agent exploration capability are integrated, a screening algorithm is matched, and multiple fault scenes are provided; a complete practical training system is built, multi-dimensional data acquisition is achieved, practical operation and evaluation iteration are standardized, the practical training quality is improved through closed-loop design, and the military crane chassis fault practical training requirements are met.

Description

Technical Field

[0001] This invention relates to the field of military equipment training technology, specifically to a military crane chassis fault reproduction and training system based on working condition simulation. Background Technology

[0002] As a key piece of equipment in military operations, military cranes integrate core subsystems such as transmission, travel, steering, and braking in their chassis, directly determining the equipment's operational stability and mission completion capabilities. The fault diagnosis and troubleshooting skills of maintenance personnel are a crucial foundation for ensuring the equipment's combat effectiveness. With the continuous upgrading of military crane technology, the chassis system structure is becoming increasingly complex, and fault types are exhibiting diverse and complex characteristics. Traditional theoretical teaching can no longer meet the needs of training maintenance personnel's practical skills. Hands-on training has become the core approach to improving skills. To ensure that maintenance personnel can cope with various chassis faults in real battlefield environments, it is necessary to construct training scenarios that closely resemble actual working conditions. Through fault reproduction, trainees can accumulate practical experience. Fault reproduction training systems based on working condition simulation can recreate the real-world scenarios and characteristics of chassis faults, helping trainees establish a connection between theoretical knowledge and practical skills. Therefore, this has become an important development direction in the field of military crane maintenance training. Its core requirement is to achieve the realism and diversity of fault scenarios and the scientific and effective nature of the training process to match the high standards required for military equipment maintenance and support.

[0003] Traditional training technologies for military crane chassis have many limitations and cannot meet the needs of modern military maintenance training. Traditional training systems rely heavily on pre-set fixed fault cases, and the types of faults are limited to typical faults of a single system. They lack coverage of complex faults involving multiple components and systems, and fail to expose trainees to diverse and complex fault scenarios that may occur in actual operations. This results in trainees having insufficient ability to handle real faults after training. At the same time, traditional systems lack a systematic screening mechanism for the teaching value of faults. Some fault cases are not well-suited to the training objectives, making it difficult to effectively improve trainees' core skills. In addition, the evaluation methods of traditional training are relatively simple, focusing on the evaluation of troubleshooting results and neglecting the assessment of key dimensions such as operational standardization and emergency handling capabilities. Furthermore, they lack dynamic optimization mechanisms, and the fault database cannot be updated according to trainees' training performance and actual needs. The management of training tools lacks standardized guidance, which can easily lead to non-standard operation and further affect the quality of training, making it difficult to adapt to the ever-evolving needs of military crane chassis maintenance training. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a military crane chassis fault reproduction and training system based on working condition simulation. It uses real vehicle components to build a physical platform, preserving the actual mechanical and power relationships. A dual-mode fault generation mechanism provides basic faults and emergent faults explored autonomously by the intelligent agent. Multiple sensors comprehensively collect chassis status, intelligent agent actions, and trainee operation data. Trainees conduct standardized troubleshooting in the practical unit, and the training evaluation and iteration unit generates multi-dimensional reports based on the data. Algorithms optimize the fault database and intelligent agent exploration strategies to form a closed-loop optimization. This system achieves realistic training scenarios, diversified fault types, and standardized training processes, providing efficient training support for military crane maintenance personnel and helping to improve their skills.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a military crane chassis fault reproduction and training system based on working condition simulation, the system comprising:

[0006] Solid chassis assembly unit: Select physical components of the transmission, travel, steering and braking subsystems of military truck cranes, install them on a high-strength structural frame according to the layout of the actual vehicle, retain the mechanical connection and power transmission relationship, pre-set standardized interfaces, and define the safety threshold of components and the standard parameter benchmark values ​​of the system;

[0007] Dual-mode fault generation unit: Based on the structure and parameters of the unit built on the physical chassis, an initial fault library is constructed, and a fault exploration agent equipped with a fault teaching value screening algorithm is embedded. The agent autonomously explores the combination of fault simulation component actions, generates emergent faults and stores them in the advanced fault library. It is equipped with a safety constraint component, and when a fault is triggered, the data acquisition process is started simultaneously.

[0008] Multi-dimensional data acquisition unit: Responding to the trigger signal of the dual-mode fault generation unit, it collects chassis status data, fault exploration agent action data and trainee troubleshooting operation data through various sensors deployed at key measurement points of the chassis. After standardized processing, the data is transmitted to the fault diagnosis practice unit and the training evaluation and fault iteration unit.

[0009] Fault Diagnosis Practice Unit: Based on the physical structure of the unit built on the solid chassis, it receives the fault scenarios provided by the dual-mode fault generation unit. Students combine the data transmitted by the multi-dimensional data acquisition unit, select the matching testing tools to carry out data verification, point investigation and fault troubleshooting operations, and the operation data is fed back to the multi-dimensional data acquisition unit in real time.

[0010] Training Evaluation and Fault Iteration Unit: Based on the full data from the multi-dimensional data acquisition unit, a training evaluation report for trainees is generated. The report is analyzed using a dynamic iterative optimization algorithm based on the fault database, and optimization suggestions are generated and fed back to the dual-mode fault generation unit to improve the fault exploration agent.

[0011] Furthermore, the high-strength structural frame of the physical chassis assembly unit is welded with anti-slip and shock-absorbing pads at the bottom, and maintenance windows are reserved on the sides, each equipped with a quickly removable aluminum alloy cover plate; the preset mechanical interfaces are distributed on the chassis side beams, with a total of twenty-four M16 to M24 threaded interfaces, including twelve M16 interfaces, eight M18 interfaces, and four M24 interfaces, each equipped with a dust plug and a label; the preset electrical interfaces include a CAN bus interface, a 12V DC power interface, and a 24V DC power interface, integrated in the electrical junction box of the chassis side beams; the preset data interfaces include an Ethernet interface and a USB expansion interface; the component safety thresholds include a maximum clutch torque of 500 Nm, a maximum driveshaft speed of 2000 rpm, a maximum steering tie rod adjustment travel of ±15 mm, a maximum brake disc operating temperature of 250 degrees Celsius, and a maximum suspension compression of 80 mm; the system standard parameter benchmark values ​​include a standard brake line pressure of 10 MPa, a standard steering system hydraulic pressure of 8 MPa, a normal operating temperature of 80 degrees Celsius for the transmission, and a standard tire pressure of 0.8 MPa.

[0012] Furthermore, the initial fault database of the dual-mode fault generation unit covers typical single faults and simple combination faults of the four major systems of military truck cranes: transmission, driving, steering, and braking. The transmission system includes clutch slippage, driveshaft imbalance, and gearbox noise; the driving system includes suspension failure, tire pressure loss, and shock absorber oil leakage; the steering system includes heavy steering, steering deviation, and steering noise; and the braking system includes brake fluid leakage, brake disc wear, and ABS sensor failure. Each fault in the initial fault database is based on the system standard parameter benchmark value defined by the physical chassis construction unit, with three preset levels of fault parameter deviation. The scope is defined for each fault entry, specifying the corresponding simulated component, triggering method, and parameter adjustment path. The advanced fault library specifically stores emergent faults generated by the fault exploration agent. Each emergent fault entry includes the action combination sequence of the fault simulation component, the chassis state characteristics under the fault condition, the fault triggering conditions, and the key points for preliminary troubleshooting. The advanced fault library is divided into three subdirectories according to the number of systems involved in the fault: single system, dual system, and multi system. Each subdirectory can store up to fifteen fault entries. Fault entries are numbered in order of generation time. Newly generated emergent faults are stored in the corresponding subdirectory after being judged to meet the criteria by the fault teaching value screening algorithm.

[0013] Furthermore, the operation flow of the fault exploration agent in the dual-mode fault generation unit is as follows: The agent reads the component safety threshold and system standard parameter benchmark values ​​defined by the physical chassis construction unit, and simultaneously retrieves the fault types and parameter ranges in the initial fault database to construct an action space containing independent actions of fault simulation components and combined actions of multiple components, clarifying the constraints for action execution; The agent randomly selects action combinations from the action space, sends execution instructions to the fault simulation components, and drives the components to act according to the instructions to generate fault conditions; The agent receives chassis status data and student troubleshooting operation data transmitted by the multi-dimensional data acquisition unit, calls the fault teaching value screening algorithm to calculate the teaching value index of the corresponding fault, and determines whether the fault meets the criteria for storage in the advanced fault database; The agent statistically analyzes the teaching value index results corresponding to different action combinations, increases the exploration weight of action combinations with high teaching value indices, decreases the exploration weight of action combinations with low teaching value indices, and updates the action space and exploration strategy in conjunction with optimization suggestions from the training evaluation and fault iteration unit.

[0014] Furthermore, in the dual-mode fault generation unit, the mathematical expression for the fault teaching value screening algorithm is: ,in The fault complexity coefficient is... Provide feedback on effectiveness to trainees. For safety and compliance coefficients, This is an increase in the concealment of faults. For characteristic reference value; when : Determined to be an intermediate-level teaching value fault, stored in the intermediate-level directory of the advanced fault database; when : If determined to be a high-level teaching value fault, it is stored in the high-level fault database directory; when Abandon the faulty combination and start exploring again.

[0015] Furthermore, the sensors in the multi-dimensional data acquisition unit include diffused silicon pressure sensors, displacement sensors, torque sensors, temperature sensors, and capacitive microphone noise sensors. There are eight diffused silicon pressure sensors, installed at two outlets of the master cylinder in the brake line, two outlets of the hydraulic pump in the steering system, one inlet of the transmission oil line, one outlet of the transmission oil line, and two in the clutch hydraulic line. There are six displacement sensors, installed at two steering tie rod adjustment ends, two suspension piston rods, and two at the brake pedal travel point. There is one torque sensor, installed in series in the middle section of the drive shaft. There are eight temperature sensors, fitted to the surface of the transmission housing and brake disc. There are six noise sensors, installed at two above the transmission, one on each side of the braking system, and two near the steering gear. The acquisition frequency of all sensors is uniformly set to 100 Hz. The acquired raw data is processed by a first-order RC low-pass filter, converted into digital signals, and encapsulated in JSON format. The encapsulation fields include data type, acquisition timestamp, value, unit, and sensor number.

[0016] Furthermore, the testing tools for the fault diagnosis practical unit are stored in an ergonomic toolbox on the side of the test bench. The toolbox is divided into sections for pressure measurement tools, torque measurement tools, electrical testing tools, and mechanical repair tools. Each tool is equipped with a dedicated slot and an RFID tag reader. The system issues an audible and visual warning when a tool is not returned to its proper place. The toolbox door is equipped with an electromagnetic lock, which automatically unlocks when the training begins and automatically locks when the training ends. After troubleshooting, the trainee presses the test start button on the test bench, and the system starts the chassis simulation operation. The running time is set to three minutes, during which key parameters such as braking pressure, steering displacement, and drive shaft torque are collected in real time. If the parameter fluctuation within three minutes does not exceed ±5% of the standard value, the fault is considered to have been successfully troubleshooted.

[0017] Furthermore, in the training evaluation and fault iteration unit, the student training evaluation report is generated according to the training evaluation indicator scoring standard. The training evaluation indicator scoring standard is 30 points for operational standardization, 40 points for fault location accuracy, 20 points for troubleshooting efficiency, and 10 points for emergency handling ability. Among them, the operational standardization score includes 10 points for correct tool use, 10 points for compliant operation steps, 5 points for proper tool storage, and 5 points for safe operation; the fault location accuracy score includes 20 points for accurately locating the fault point, 10 points for no missed fault points, and 10 points for no misjudged fault points; the troubleshooting efficiency score is set according to the standard time consumption, and full marks are obtained for completing within the standard time consumption, and two points are deducted for each minute exceeding the standard time consumption; the emergency handling ability score is based on the correct response to emergencies as the full mark standard; the fault database dynamic iteration optimization cycle is set to automatically start once after every 30 training sessions, and each iteration extracts the historical TVI average value of faults, troubleshooting data of no less than 20 students, and instructor teaching suitability scores.

[0018] Furthermore, in the training evaluation and fault iteration unit, the mathematical expression of the fault database dynamic iterative optimization algorithm is: ,in, For fault iteration optimization coefficients, As a teaching value index, For population fitness coefficient, This is a stability correction factor. The standard deviation of troubleshooting time; when : Fault retention, maintaining original parameters; when : Optimize the fault parameters and re-store them in the fault database; when Remove from the fault database to free up storage resources.

[0019] Furthermore, in the training evaluation and fault iteration unit, the process of forming optimization suggestions is as follows: Extract all data transmitted from the multi-dimensional data acquisition unit, classify and organize it according to student operation data, fault condition data, and fault exploration agent action data, remove invalid data, and complete data association matching; analyze student training performance from the dimensions of operational standardization, fault location accuracy, troubleshooting efficiency, and emergency handling capability, statistically analyze common problems in student operations corresponding to different fault types, and simultaneously analyze the call frequency and student completion status of each emergent fault in the advanced fault library; combine the FOC value output by the fault library dynamic iterative optimization algorithm to determine the conclusion of retaining, optimizing, or removing each fault case; for fault cases with FOC values ​​in the middle range, locate the specific parameters causing insufficient adaptability; form targeted optimization suggestions based on the analysis conclusions, including adjusting the action combination exploration weight of the fault exploration agent, optimizing the fault parameter deviation range, adding or deleting fault cases in the advanced fault library, and adjusting fault triggering conditions. After all optimization suggestions are classified and organized according to fault system type, they are transmitted back to the dual-mode fault generation unit to improve the action space and exploration strategy of the fault exploration agent.

[0020] Compared with existing technologies, this military crane chassis fault reproduction and training system based on working condition simulation has the following advantages:

[0021] I. This invention constructs a dual-mode fault generation unit, integrating the autonomous exploration capabilities of an initial fault database and a fault-exploring intelligent agent to achieve diversified fault scenario supply. The initial fault database covers typical faults in transmission, driving, steering, and braking systems, providing stable scenarios for basic training. The fault-exploring intelligent agent constructs an action space based on system structure and parameters, autonomously explores multi-component combined actions, and generates emergent faults with practical teaching value. Combined with a fault teaching value screening algorithm, the generated faults are scientifically screened to ensure that the faults stored in the advanced fault database are adapted to the training needs of different stages. At the same time, the safety constraint component simultaneously starts data acquisition to ensure the safety and controllability of the training process. This design breaks through the limitations of traditional training fault types being singular and lacking complexity, allowing trainees to be exposed to various fault scenarios from basic to advanced, effectively improving their ability to diagnose, troubleshoot, and eliminate different faults, and meeting the actual teaching needs of military crane chassis fault training.

[0022] Second, this invention establishes a complete system for multi-dimensional data acquisition, fault diagnosis practice, and iterative training evaluation, achieving standardization, datafication, and continuous optimization of the training process. The multi-dimensional data acquisition unit comprehensively captures chassis status, agent actions, and student operation data, providing sufficient basis for training evaluation. The fault diagnosis practice unit is equipped with partitioned storage tools and standardized fault verification processes to ensure the standardization and accuracy of practical training. The training evaluation and fault iteration unit generates multi-dimensional evaluation reports based on full data, analyzes fault adaptability through dynamic iterative optimization algorithms, and forms targeted optimization suggestions that are fed back to the fault generation unit to continuously improve the action space and exploration strategy of the fault-exploring agent. This closed-loop design ensures the standardization of training operations and the objectivity of evaluation, while continuously optimizing training content based on student performance, ensuring that the training system always adapts to teaching needs and significantly improving training quality and students' comprehensive practical skills.

[0023] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0025] Figure 1 System flowchart for military crane chassis training system;

[0026] Figure 2 System framework diagram of a military crane chassis training system;

[0027] Figure 3 Flowchart of the multi-dimensional data acquisition unit for a military crane chassis training system;

[0028] Figure 4 This is a structural diagram of the training platform for a military crane chassis training system.

[0029] Figure 5 Detailed diagram of the training platform for a military crane chassis training system. Detailed Implementation

[0030] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0031] Example 1:

[0032] Basic troubleshooting training scenario for beginner trainees (single fault in transmission system and braking system).

[0033] This scenario is designed for beginners new to military crane chassis repair. It focuses on typical single faults in the transmission and braking systems, aiming to help trainees familiarize themselves with the basic chassis structure, master the use of basic testing tools, and understand the troubleshooting logic for single faults. The training lasts 4 class hours, with 2 trainees per group, for a total of 15 group training sessions. Specific steps are as follows: Figure 1 As shown.

[0034] The physical chassis assembly unit was prepared in advance for practical training. Its high-strength structural frame features anti-slip and shock-absorbing feet welded to the bottom, ensuring stability during training. The side inspection windows are equipped with quick-release aluminum alloy covers, allowing trainees to directly observe the internal component structure and connections. The chassis side beams have 24 pre-installed threaded interfaces: twelve M16, eight M18, and four M24, all equipped with dust plugs and labels, allowing trainees to quickly identify interface types and uses. The electrical junction box integrates a CAN bus interface, 12V and 24V DC power interfaces, and data interfaces including an Ethernet interface and a USB expansion interface, all in normal working order, ensuring stable data transmission and equipment power supply during training. Furthermore, the unit has clearly defined component safety thresholds, including a maximum clutch torque of 500 Nm and a maximum brake disc operating temperature of 250 degrees Celsius, as well as system standard parameter benchmarks, such as a standard brake line pressure of 10 MPa and a normal gearbox operating temperature of 80 degrees Celsius, providing trainees with accurate parameter references and helping them establish a understanding of standard parameters. Figure 4 As shown.

[0035] After the training begins, the dual-mode fault generation unit selects clutch slippage in the transmission system and brake fluid leakage in the braking system from the initial fault database based on the training objectives of the beginner trainees. Both faults are typical single faults, and their deviation from the preset three-level parameters is set as a level one minor fault, allowing trainees to gradually master basic fault troubleshooting methods. The unit's built-in safety constraint component is activated simultaneously to ensure that the fault triggering process complies with safety regulations, avoiding component damage or safety hazards during training. At the same time, it triggers the synchronous data acquisition process of the multi-dimensional data acquisition unit, providing complete data support for subsequent trainee operation evaluation and fault analysis.

[0036] After responding to the trigger signal, the multi-dimensional data acquisition unit begins to comprehensively collect data through sensors deployed at key locations on the chassis, ensuring trainees can obtain accurate fault diagnosis data. Two diffused silicon pressure sensors are installed at the master cylinder outlet of the brake line to collect brake fluid pressure data in real time, helping trainees determine brake fluid leakage through pressure changes; a torque sensor is installed in series in the middle of the drive shaft to collect clutch transmission torque data, assisting trainees in identifying clutch slippage faults; in addition, a temperature sensor mounted on the gearbox housing and noise sensors installed on both sides of the braking system, totaling twenty-nine sensors (eight diffused silicon pressure sensors, six displacement sensors, one torque sensor, eight temperature sensors, and six noise sensors), all collect data at a frequency of 100 Hz, ensuring the continuity and timeliness of data acquisition. All raw data is processed by a first-order RC low-pass filter, converted into digital signals, and encapsulated in JSON format. The encapsulation fields include data type, acquisition timestamp, numerical unit, and sensor number, making the data clear and standardized, easy for trainees to read and analyze, and then transmitted to the fault diagnosis practice unit and the training evaluation and fault iteration unit, such as... Figure 3 As shown.

[0037] Trainees entered the fault diagnosis practical exercise, where they obtained fault scenario prompts from the dual-mode fault generation unit and real-time data transmitted from the multi-dimensional data acquisition unit through the fault diagnosis practical exercise unit, quickly clarifying the direction of troubleshooting. They retrieved corresponding pressure measuring tools and mechanical repair tools from the ergonomic toolbox on the side of the bench. The toolbox was functionally divided, and each tool was equipped with a dedicated slot and RFID tag reader, guiding trainees to develop standardized habits of tool handling and storage. Trainees first diagnosed a brake fluid leak based on abnormal brake line pressure data, checked the sealing condition of the brake line interface through the inspection window, and replaced any damaged seals. Next, they diagnosed a clutch slippage fault based on fluctuations in driveshaft torque data and the clutch's working status, adjusting the clutch clearance. After troubleshooting, trainees pressed the test start button on the bench, and the system started a three-minute chassis simulation run, during which key parameters such as brake pressure and driveshaft torque were collected in real time. Ultimately, because the parameter fluctuation within three minutes did not exceed ±5% of the standard value, the fault was considered successfully resolved, directly verifying the trainee's troubleshooting effectiveness. Throughout the entire operation, data such as tool usage, operation steps, and troubleshooting time were fed back to the multi-dimensional data acquisition unit in real time, comprehensively recording the trainee's practical performance.

[0038] The training evaluation and fault iteration unit receives all data transmitted from the multi-dimensional data acquisition unit and generates a student training evaluation report according to the training evaluation index scoring standard, clearly presenting the student's ability level in each dimension. Regarding operational standardization, a perfect score of 30 points is awarded for correct tool use, compliant operation procedures, proper tool storage after training, and no safety issues. For fault location accuracy, a perfect score of 40 points is awarded for accurately locating two fault points without omissions or misjudgments. For troubleshooting efficiency, a perfect score of 20 points is awarded for completing troubleshooting within the standard time limit. For emergency handling capability, a perfect score of 10 points is awarded for no unexpected situations. The overall score is 100 points, allowing students and instructors to quickly understand the training results. Simultaneously, the unit initiates a dynamic iterative optimization algorithm for the fault database after every 30 training sessions. The mathematical expression for the dynamic iterative optimization algorithm for the fault database is: ,in, For fault iteration optimization coefficients, As a teaching value index, For population fitness coefficient, This is a stability correction factor. The standard deviation of troubleshooting time; when : Fault retention, maintaining original parameters; when : Optimize the fault parameters and re-store them in the fault database; when Removed from the fault database to free up storage resources. The current fifteen sets of training data have not yet reached the iteration cycle. Only statistical analysis of the training data in this session is performed to form a preliminary conclusion that the basic trainees have good adaptability to troubleshooting this type of single fault and can maintain the existing fault parameters. After subsequent training data is supplemented, formal iteration and optimization will be completed. Optimization suggestions will be fed back to the dual-mode fault generation unit to continuously improve the relevance of subsequent training.

[0039] In summary, this embodiment addresses the basic troubleshooting needs of beginner trainees, relying on the collaborative operation of various system units to complete the training. The physical chassis construction unit provides a stable physical foundation and standard parameter benchmarks, while the dual-mode fault generation unit selects typical single faults to ensure a gradual learning process. The multi-dimensional data acquisition unit accurately captures fault data, providing reliable data for trainees' troubleshooting. The fault diagnosis practice unit helps trainees use tools correctly and complete troubleshooting verification, while the training evaluation and fault iteration unit generates detailed reports, intuitively presenting the trainees' mastery of basic skills. The entire process helps beginner trainees familiarize themselves with the chassis structure and master basic troubleshooting logic, while also accumulating initial data for subsequent system iterations, fully adapting to the training objectives of the beginner stage.

[0040] Example 2:

[0041] Intermediate trainees' comprehensive troubleshooting training scenario for complex faults (combined faults in steering and driving systems + emergent faults).

[0042] This scenario is designed for intermediate learners with basic troubleshooting skills. It focuses on combined faults in the steering and driving systems, as well as emergent faults generated by fault-finding agents. The aim is to enhance learners' analytical abilities regarding multi-system interconnected faults and their skills in handling complex faults. The training lasts 6 hours, with 3 learners per group, for a total of 20 training sessions. Specific steps are as follows: Figure 2 As shown.

[0043] The physical chassis assembly unit remains in a state of readiness. All interfaces of the high-strength structural frame, including threaded interfaces, electrical interfaces, data interfaces, component safety thresholds, and system standard parameter benchmarks, meet the training requirements. Key parameters such as the maximum steering tie rod adjustment travel of ±15 mm, the maximum suspension compression of 80 mm, and the standard tire pressure of 0.8 MPa have been calibrated to ensure the accuracy of fault simulation. The aluminum alloy cover of the inspection window is removable, allowing trainees to conduct in-depth inspections of the steering and driving system components, facilitating troubleshooting of complex faults. Figure 5 As shown.

[0044] After the dual-mode fault generation unit is activated, it first retrieves a combined fault from the initial fault database: a heavy steering fault in the steering system and a suspension failure fault in the driving system. This is preset to a level two moderate fault parameter deviation range, aligning with the skill level of intermediate learners. Then, the fault exploration agent is activated. The agent first reads the component safety thresholds and system standard parameter benchmarks from the physical chassis assembly unit, retrieves the fault types and parameter ranges of the two systems from the initial fault database, constructs an action space including independent actions such as steering tie rod adjustment and suspension piston rod movement, as well as combined actions of multiple components, and clarifies the safety constraints for action execution to ensure the safety and rationality of the fault generation process. The agent randomly selects a combined action of steering tie rod maximum travel limit + suspension over-compression from the action space, sends execution commands to the fault simulation component, generating an emergent fault to train learners' ability to handle complex and unknown faults. After receiving chassis status data transmitted from the multi-dimensional data acquisition unit, the agent calls the fault teaching value screening algorithm to calculate the teaching value index of the fault. The mathematical expression of the fault teaching value screening algorithm is: ,in The fault complexity coefficient is... Provide feedback on effectiveness to trainees. For safety and compliance coefficients, This is an increase in the concealment of faults. For characteristic reference value; when : Determined to be an intermediate-level teaching value fault, stored in the intermediate-level directory of the advanced fault database; when : If determined to be a high-level teaching value fault, it is stored in the high-level fault database directory; when The fault combination was abandoned, and a new investigation was conducted. It was determined that the fault met the intermediate-level teaching value fault criteria, ensuring that the fault had sufficient educational significance. The fault was then organized according to its action combination sequence, state characteristics, triggering conditions, and preliminary troubleshooting points, and stored in the advanced fault database's dual-system subdirectory. If the current subdirectory did not contain fifteen faults, they were stored sequentially by generation time for easy retrieval and management. Simultaneously, the safety constraint component monitored the entire process to ensure that the fault generation process did not exceed the component's safety threshold, protecting the training equipment.

[0045] The multi-dimensional data acquisition unit synchronously responds to the two fault trigger signals from the dual-mode fault generation unit. Deployed sensors comprehensively collect data, covering various characteristic parameters of multiple system faults. Two of the six displacement sensors are installed at the steering tie rod adjustment end to collect steering displacement data, and two are installed on the suspension piston rod to collect suspension compression data, accurately capturing key changes in steering and driving systems. Of the eight temperature sensors, those mounted against the steering gear and gearbox housing collect temperature data. Of the six noise sensors, two are installed near the steering gear and two are installed above the gearbox to collect abnormal noise data, assisting trainees in diagnosing faults through temperature and sound anomalies. Eight diffused silicon pressure sensors collect pressure data from various oil circuits at preset locations, and one torque sensor collects driveshaft torque data, comprehensively supporting the integrated analysis of complex faults. All sensors collect raw data at a frequency of 100 Hz, which is then converted into digital signals after first-order RC low-pass filtering. The relevant fields are encapsulated in JSON format and transmitted to the fault diagnosis practice unit and the training evaluation and fault iteration unit, respectively. Simultaneously, the action data of the fault exploration agent is transmitted, providing complete data support for trainees' troubleshooting and system evaluation.

[0046] After receiving a fault diagnosis task, trainees quickly outlined their troubleshooting approach by combining the emergent fault prompts from the combined fault scenarios provided by the fault diagnosis practice unit with real-time data from the multi-dimensional data acquisition unit. They utilized a variety of tools from their ergonomic toolbox, including pressure measuring tools, torque measuring tools, electrical testing tools, and mechanical repair tools, to meet the diverse needs of complex fault diagnosis. Trainees first checked the steering system hydraulic pump pressure (standard 8 MPa) based on steering displacement data and heavy steering, discovering insufficient hydraulic pump supply and replacing the hydraulic filter. Next, based on suspension compression data and abnormal driving stability, they determined the suspension failure was due to spring elasticity decay and replaced the suspension springs. For emergent faults, combining the correlation data and fault prompts between steering and suspension, they adjusted the steering tie rod travel limit parameters to repair the linkage effect of excessive suspension compression, improving their ability to analyze and handle multi-component interconnected faults. After troubleshooting, trainees pressed the test start button, and the system ran a three-minute simulation, monitoring key parameters such as steering displacement, suspension compression, and braking pressure in real time. Ultimately, because the parameter fluctuations did not exceed ±5% of the standard values, the fault was considered successfully resolved, effectively verifying the trainees' comprehensive troubleshooting capabilities. Throughout the entire operation, data such as the trainee's tool selection, operation steps, troubleshooting sequence, time consumption, and emergency handling actions are fed back to the multi-dimensional data acquisition unit in real time, comprehensively capturing the trainee's practical performance in complex scenarios.

[0047] After receiving all the data, the training evaluation and fault iteration unit generates a report according to the evaluation indicators, accurately reflecting the trainees' ability to handle complex faults. Regarding operational standardization, two points are deducted for each tool usage error, resulting in a score of 28. For fault location accuracy, accurately locating three fault points but missing one related impact point of an emerging fault results in a deduction of five points, resulting in a score of 35. Regarding troubleshooting efficiency, exceeding the standard by two minutes results in a deduction of four points, resulting in a score of 16. Regarding emergency response capabilities, successfully handling the sudden situation of an emerging fault earns ten points, for a total score of 89, providing a basis for instructors to provide targeted guidance. Simultaneously, the unit extracts all data from this and the previous nineteen training sessions, reaching an iteration cycle of thirty training sessions. It then initiates a dynamic iterative optimization algorithm for the fault database for analysis, ensuring the fault database continuously adapts to teaching needs. The mathematical expression for the dynamic iterative optimization algorithm for the fault database is: ,in, For fault iteration optimization coefficients, As a teaching value index, For population fitness coefficient, This is a stability correction factor. The standard deviation of troubleshooting time; when : Fault retention, maintaining original parameters; when : Optimize the fault parameters and re-store them in the fault database; when The algorithm removes faults from the fault database, freeing up storage resources. It combines the average historical teaching value index of faults with troubleshooting data from twenty trainees and instructor teaching suitability scores to calculate the fault iteration optimization coefficient for each fault. For combined faults in the initial fault database, those with iteration optimization coefficients between 1.5 and 3 are optimized and re-stored to improve fault suitability. For emergent faults, those with iteration optimization coefficients greater than or equal to 3 are retained to enrich the complex fault case database. Simultaneously, statistics show that steering and suspension linkage action combinations have high teaching value indices, leading to optimization suggestions for adjusting the exploration weights of the fault exploration agent and optimizing the fault parameter range. These suggestions are categorized by system type and then transmitted back to the dual-mode fault generation unit to improve the action space and exploration strategy of the fault exploration agent, making subsequent training more aligned with the skill enhancement needs of intermediate trainees.

[0048] In summary, this embodiment focuses on enhancing intermediate learners' ability to handle complex faults, with efficient collaboration among all system units. The physical chassis construction unit calibrates key parameters to support multi-system fault simulation. The dual-mode fault generation unit combines initial library-based faults with emergent faults generated by fault exploration agents, and a fault teaching value screening algorithm ensures its educational significance. The multi-dimensional data acquisition unit comprehensively captures multi-system fault data, assisting learners in comprehensive analysis. The fault diagnosis practice unit trains learners' complex troubleshooting skills, while the training evaluation and fault iteration unit optimizes fault cases and exploration strategies through a dynamic iterative optimization algorithm based on the fault library. The training not only improves learners' ability to handle multi-system interconnected faults but also drives the system to continuously adapt to the teaching needs of the intermediate level, achieving a two-way empowerment of skill enhancement and system optimization.

[0049] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A military crane chassis fault reproduction and training system based on working condition simulation, characterized in that, The system includes: Solid chassis assembly unit: Select physical components of the transmission, travel, steering and braking subsystems of military truck cranes, install them on a high-strength structural frame according to the layout of the actual vehicle, retain the mechanical connection and power transmission relationship, pre-set standardized interfaces, and define the safety threshold of components and the standard parameter benchmark values ​​of the system; Dual-mode fault generation unit: Based on the structure and parameters of the unit built on the physical chassis, an initial fault library is constructed, and a fault exploration agent equipped with a fault teaching value screening algorithm is embedded. The agent autonomously explores the combination of fault simulation component actions, generates emergent faults and stores them in the advanced fault library. It is equipped with a safety constraint component, and when a fault is triggered, the data acquisition process is started simultaneously. Multi-dimensional data acquisition unit: Responding to the trigger signal of the dual-mode fault generation unit, it collects chassis status data, fault exploration agent action data and trainee troubleshooting operation data through various sensors deployed at key measurement points of the chassis. After standardized processing, the data is transmitted to the fault diagnosis practice unit and the training evaluation and fault iteration unit. Fault Diagnosis Practice Unit: Based on the physical structure of the unit built on the solid chassis, it receives the fault scenarios provided by the dual-mode fault generation unit. Students combine the data transmitted by the multi-dimensional data acquisition unit, select the matching testing tools to carry out data verification, point investigation and fault troubleshooting operations, and the operation data is fed back to the multi-dimensional data acquisition unit in real time. Training Evaluation and Fault Iteration Unit: Based on the full data from the multi-dimensional data acquisition unit, a training evaluation report for trainees is generated. The report is analyzed using a dynamic iterative optimization algorithm based on the fault database, and optimization suggestions are generated and fed back to the dual-mode fault generation unit to improve the fault exploration agent.

2. The military crane chassis fault reproduction and training system based on working condition simulation according to claim 1, characterized in that, The high-strength structural frame of the solid chassis assembly unit has anti-slip and shock-absorbing pads welded to the bottom, and maintenance windows are reserved on the sides, each equipped with a quick-removable aluminum alloy cover. Pre-set mechanical interfaces are distributed on the chassis side beams, with a total of twenty-four threaded interfaces ranging from M16 to M24, including twelve M16 interfaces, eight M18 interfaces, and four M24 interfaces. Each interface is equipped with a dust plug and a label. Pre-set electrical interfaces include a CAN bus interface, a 12V DC power interface, and a 24V DC power interface, integrated into the electrical junction box on the chassis side beams. Pre-set data interfaces include an Ethernet interface and a USB expansion interface. Component safety thresholds include a maximum clutch torque of 500 Nm, a maximum driveshaft speed of 2000 rpm, a maximum steering tie rod adjustment travel of ±15 mm, a maximum brake disc operating temperature of 250 degrees Celsius, and a maximum suspension compression of 80 mm. System standard parameter benchmarks include a standard brake line pressure of 10 MPa, a standard steering system hydraulic pressure of 8 MPa, a normal gearbox operating temperature of 80 degrees Celsius, and a standard tire pressure of 0.8 MPa.

3. The military crane chassis fault reproduction and training system based on working condition simulation according to claim 1, characterized in that, The initial fault database of the dual-mode fault generation unit covers typical single faults and simple combination faults of the four major systems of military truck cranes: transmission, driving, steering, and braking. The transmission system includes clutch slippage, drive shaft imbalance, and abnormal gearbox noise; the driving system includes suspension failure, tire pressure loss, and shock absorber oil leakage; the steering system includes heavy steering, steering deviation, and abnormal steering noise; and the braking system includes brake fluid leakage, brake disc wear, and ABS sensor failure. Each fault in the initial fault database is based on the system standard parameter benchmark value defined by the physical chassis construction unit, with three preset levels of fault parameter deviation ranges. Each fault entry clearly defines the corresponding simulated component, triggering method, and parameter adjustment path. The advanced fault database is specifically designed to store emergent faults generated by the fault exploration agent. Each emergent fault entry includes the action combination sequence of the fault simulation component, the chassis state characteristics under the fault condition, the fault triggering conditions, and the key points for preliminary troubleshooting. The advanced fault database is divided into three subdirectories based on the number of systems involved in the fault: single system, dual system, and multi system. Each subdirectory can store up to fifteen fault entries. Fault entries are numbered in chronological order of their generation. Newly generated emergent faults are stored in the corresponding subdirectory after being judged to meet the criteria by the fault teaching value screening algorithm.

4. The military crane chassis fault reproduction and training system based on working condition simulation according to claim 1, characterized in that, The operation flow of the fault exploration agent in the dual-mode fault generation unit is as follows: the agent reads the component safety threshold and system standard parameter benchmark value defined by the physical chassis construction unit, and at the same time retrieves the fault type and parameter range in the initial fault database to construct an action space that includes independent actions of fault simulation components and combined actions of multiple components, and clarifies the constraints for action execution; the agent randomly selects action combinations from the action space, sends execution instructions to the fault simulation components, and drives the components to act according to the instructions to generate fault conditions. The agent receives chassis status data and student troubleshooting operation data transmitted by the multi-dimensional data acquisition unit, calls the fault teaching value screening algorithm to calculate the teaching value index of the corresponding fault, and determines whether the fault meets the criteria for storage in the advanced fault database. The agent counts the teaching value index results corresponding to different action combinations, increases the exploration weight of action combinations with high teaching value index, and decreases the exploration weight of action combinations with low teaching value index. At the same time, it updates the action space and exploration strategy by combining the optimization suggestions from the training evaluation and fault iteration unit feedback.

5. The military crane chassis fault reproduction and training system based on working condition simulation according to claim 1, characterized in that, In the dual-mode fault generation unit, the mathematical expression of the fault teaching value screening algorithm is: ,in The fault complexity coefficient is... Provide feedback on effectiveness to trainees. For safety and compliance coefficients, This is an increase in the concealment of faults. For characteristic reference value; when : Determined to be an intermediate-level teaching value fault, stored in the intermediate-level directory of the advanced fault database; when : If determined to be a high-level teaching value fault, it is stored in the high-level fault database directory; when Abandon the faulty combination and explore a new one.

6. The military crane chassis fault reproduction and training system based on working condition simulation according to claim 1, characterized in that, The sensors in the multidimensional data acquisition unit include diffused silicon pressure sensors, displacement sensors, torque sensors, temperature sensors, and capacitive microphone noise sensors; there are eight diffused silicon pressure sensors in total, which are respectively installed at two outlets of the master cylinder in the brake line, two outlets of the hydraulic pump in the steering system, one inlet of the transmission oil line, one outlet of the transmission oil line, and two in the clutch hydraulic line; there are six displacement sensors in total, which are respectively installed at two ends of the steering tie rod, two on the suspension piston rod, and two at the brake pedal travel; and one torque sensor is installed in series in the middle section of the drive shaft; There are eight temperature sensors, which are attached to the gearbox housing and brake disc surface; there are six noise sensors, two of which are installed above the gearbox, one on each side of the braking system, and two near the steering gear; the acquisition frequency of all sensors is uniformly set to 100 Hz, and the acquired raw data is processed by a first-order RC low-pass filter, converted into digital signals and encapsulated in JSON format. The encapsulation fields include data type, acquisition timestamp, value, unit, and sensor number.

7. The military crane chassis fault reproduction and training system based on working condition simulation according to claim 1, characterized in that, The testing tools in the fault diagnosis practice unit are stored in an ergonomic toolbox on the side of the bench. The toolbox is divided into pressure measurement tool area, torque measurement tool area, electrical testing tool area and mechanical repair tool area. Each tool is equipped with a dedicated card slot and RFID tag reader. The system will issue an audible and visual prompt when the tool is not returned to its place. The toolbox cabinet door is equipped with an electromagnetic lock, which will automatically unlock when the training begins and automatically lock when the training ends. After troubleshooting, the trainee presses the test start button on the test bench, and the system starts the chassis simulation operation. The running time is set to three minutes. During this time, key parameters such as braking pressure, steering displacement, and drive shaft torque are collected in real time. If the parameter fluctuation range does not exceed plus or minus five percent of the standard value within three minutes, the troubleshooting is considered successful.

8. The military crane chassis fault reproduction and training system based on working condition simulation according to claim 1, characterized in that, In the training evaluation and fault iteration unit, the trainee training evaluation report is generated according to the training evaluation indicator scoring standard. The training evaluation indicator scoring standard is 30 points for operational standardization, 40 points for fault location accuracy, 20 points for troubleshooting efficiency, and 10 points for emergency handling ability. Among them, the operational standardization score includes 10 points for correct tool use, 10 points for compliant operation steps, 5 points for proper tool storage, and 5 points for safe operation; the fault location accuracy score includes 20 points for accurately locating the fault point, 10 points for no missed fault points, and 10 points for no misjudged fault points; the troubleshooting efficiency score is set according to the standard time consumption. The full score is obtained by completing the task within the standard time consumption, and 2 points are deducted for each minute exceeding the standard time consumption; the emergency handling ability score is based on the correct response to emergencies as the full score standard; the fault database dynamic iteration optimization cycle is set to automatically start once after every 30 training sessions. Each iteration extracts the historical TVI average value of the fault, troubleshooting data of no less than 20 trainees, and instructor teaching suitability score.

9. The military crane chassis fault reproduction and training system based on working condition simulation according to claim 1, characterized in that, In the training evaluation and fault iteration unit, the mathematical expression of the fault database dynamic iterative optimization algorithm is: ,in, For fault iteration optimization coefficients, As a teaching value index, For population fitness coefficient, This is a stability correction factor. The standard deviation of troubleshooting time; when : Fault retention, maintaining original parameters; when : Optimize the fault parameters and re-store them in the fault database; when Remove from the fault database to free up storage resources.

10. The military crane chassis fault reproduction and training system based on working condition simulation according to claim 1, characterized in that, In the training evaluation and fault iteration unit, the process of forming optimization suggestions is as follows: extract all data transmitted by the multi-dimensional data acquisition unit, classify and organize it according to student operation data, fault condition data, and fault exploration agent action data, remove invalid data and complete data association matching; break down student training performance from the dimensions of operation standardization, fault location accuracy, troubleshooting efficiency, and emergency handling capability, statistically analyze common problems in student operation corresponding to different fault types, and analyze the call frequency and student completion status of each emergent fault in the advanced fault library. Based on the FOC value output by the fault database dynamic iterative optimization algorithm, the conclusions for retaining, optimizing, or eliminating each fault case are determined. For fault cases with FOC values ​​in the middle range, the specific parameters causing insufficient adaptability are identified. Targeted optimization suggestions are generated based on the analysis conclusions, including adjusting the action combination exploration weights of the fault exploration agent, optimizing the deviation range of fault parameters, adding or deleting fault cases in the advanced fault database, and adjusting fault triggering conditions. All optimization suggestions are categorized and organized according to the fault system type and then transmitted back to the dual-mode fault generation unit to improve the action space and exploration strategy of the fault exploration agent.

Images