Method and system for building flight control computer simulation test system

Abstract

The present invention proposes a method and system for building a flight control computer simulation test system, which belongs to the field of aircraft technology. The method includes constructing a structured simulation test knowledge base for the aircraft; constructing a flight control semi-physical simulation test environment and an intelligent agent for operating and perceiving the flight control semi-physical simulation test environment, wherein the intelligent agent is formed by training the structured simulation test knowledge base for the aircraft; constructing a perception layer, a decision layer, a memory layer, and an execution layer for the intelligent agent to form a flight control simulation intelligent agent; and physically connecting the flight control semi-physical simulation test environment with the flight control simulation intelligent agent to form a flight control computer simulation test system. The present invention utilizes the flight control simulation intelligent agent to perform test design, test execution control, and result analysis for the flight control semi-physical simulation test environment, thereby achieving the purpose of autonomously completing simulation test experiments without relying on manpower, avoiding the influence of human error, and improving the reliability of simulation test results.

Description

Technical Field

[0001] The present invention belongs to the field of aircraft technology, and relates to a technology for simulating an aircraft's flight control computer by combining it with an intelligent agent, specifically a method and system for building a flight control computer simulation test system. Background Art

[0002] The flight control computer is a key component of an aircraft, particularly as a core module of unmanned aerial vehicles (UAVs). Its performance directly impacts the aircraft's flight safety and mission execution capabilities. Due to the high cost of aircraft design and the potential risks associated with the testing phase, simulation testing has become an essential verification step in the aircraft design process. Simulation testing of the flight control computer simulates the aircraft's actual flight environment to test and verify the computer's software and hardware, thereby improving R&D and design efficiency, effectively shortening the design cycle, and reducing both the cost and risk of aircraft design.

[0003] However, as the functions of flight control computers become more and more diverse, the types of simulation test data and simulation project requirements of flight control computer simulation test systems are also becoming more and more diverse. At the same time, the reliability requirements for test results generated by flight control computer simulation test systems are also becoming higher and higher. The test design, test execution, and result analysis processes in the flight control computer simulation test process basically need to be completed manually, which requires a large amount of manpower to be invested in flight control computer simulation testing. When processing simulation test data, it is easy to be interfered by human factors, resulting in large errors in the simulation test data processing process, affecting the reliability of the simulation test results. Summary of the Invention

[0004] In response to the technical problem described in the above background technology, the existing flight control computer simulation test system requires a large amount of manpower during the simulation test process, and is easily interfered by human factors when processing the simulation test data, resulting in large errors in the simulation test data processing process, which affects the reliability of the simulation test results. To address this technical problem, the present invention proposes a method and system for building a flight control computer simulation test system.

[0005] The present invention introduces an intelligent agent into the flight control computer simulation test, uses the structured simulation test knowledge base of the aircraft to train the intelligent agent, and uses the intelligent agent to complete the test design, test execution and result analysis of the flight control computer simulation test, thereby achieving the purpose of autonomously completing the simulation test without relying on manpower, avoiding human errors introduced in the flight control computer simulation test process, and improving the reliability of the simulation test results.

[0006] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0007] The method for constructing a flight control computer simulation test system of the present invention comprises:

[0008] Build a knowledge base for structured simulation testing of aircraft;

[0009] Build a flight control semi-physical simulation test environment and an intelligent agent that operates and perceives the flight control semi-physical simulation test environment. The intelligent agent is trained using a structured simulation test knowledge base for the aircraft.

[0010] The intelligent agent is constructed with a perception layer, a decision layer, a memory layer, and an execution layer to form a flight control simulation intelligent agent. The decision layer of the intelligent agent uses a scoring function and an optimization function to screen decision paths. Then, a dynamic pruning strategy is used to remove decision paths that do not meet the dynamic pruning conditions after screening. The dynamic pruning conditions include the decision path threshold and the number of executions. The decision path refers to the flight path of the aircraft in the simulation test.

[0011] The flight control semi-physical simulation test environment is physically connected to the flight control simulation intelligent body to form a flight control computer simulation test system.

[0012] It is further defined that the screening using the scoring function of the decision path and the optimization function of the decision path specifically includes:

[0013] The total score of each decision path is determined based on the node action of each decision path and the reward obtained after the node action is executed;

[0014] Construct a cost function, normalize the cost function, and form an optimization function for the decision path;

[0015] The total scores of all decision paths are judged according to the optimization function of the decision path, and the decision paths are screened to form the screened decision paths.

[0016] It is further defined that the decision paths that do not meet the dynamic pruning conditions after being eliminated by the dynamic pruning strategy specifically include:

[0017] A validity pruning formula for decision paths is constructed, and the screened decision paths are dynamically evaluated according to the validity pruning formula. Decision paths that do not meet the dynamic pruning conditions are eliminated from the screened decision paths based on the dynamic evaluation results.

[0018] It is further defined that the construction of a structured simulation test knowledge base for an aircraft specifically includes:

[0019] Obtain the aircraft simulation test knowledge base;

[0020] The data organization algorithm, data cleaning strategy, data segmentation algorithm and standardization processing algorithm are used to classify, organize, clean noise, segment data and label data in sequence to form a structured simulation test knowledge base of aircraft.

[0021] It is further defined that the scoring function of the decision path for:

[0022]

[0023] Where, , is the i-th decision path of the aircraft in the simulation test; is the t-th node action on the i-th decision path, t≥1; T is the total number of node actions on the i-th decision path; is the decision factor, r t is the reward score of the t-th node action; is the penalty weight parameter; is the optimization function of the i-th decision path of the aircraft in the simulation test;

[0024] The validity pruning formula is:

[0025] and

[0026] Where, is the scoring threshold of the decision path; S i is the total number of successful node actions executed by the aircraft in the i-th decision path in the simulation test; n i N is the total number of node actions executed by the aircraft in the i-th decision path in the simulation test; min The minimum valid execution count.

[0027] It is further defined that the method for building the flight control computer simulation test system also includes establishing a data dictionary, and associating the perception layer and decision layer of the intelligent body through the data dictionary, so that the decision layer of the intelligent body can call the information perceived in the perception layer of the intelligent body.

[0028] It is further defined that the method for building the flight control computer simulation test system also includes configuring a standard API calling interface, and the flight control semi-physical simulation test environment and the flight control simulation intelligent agent are physically connected through the standard API calling interface.

[0029] It is further defined that the simulation test knowledge base of the aircraft includes initial scenario information of the flight control computer, simulation test purpose information of the flight control computer, simulation test method information of the flight control computer, simulation test process information of the flight control computer, simulation test data analysis method of the flight control computer, simulation test result judgment criteria of the flight control computer, sensor data of the flight control computer, simulation case analysis of the flight control computer, simulation test outline of the flight control computer and simulation test report of the flight control computer.

[0030] The flight control computer simulation test system construction system formed based on the above-mentioned flight control computer simulation test system construction method includes:

[0031] Building module: used to build a structured simulation test knowledge base for aircraft;

[0032] The first building module is used to build a flight control semi-physical simulation test environment and an intelligent agent that operates and perceives the flight control semi-physical simulation test environment. The intelligent agent is trained using a structured simulation test knowledge base for the aircraft.

[0033] The second building module is used to build the perception layer, decision layer, memory layer, and execution layer for the intelligent agent to form a flight control simulation intelligent agent. The decision layer of the intelligent agent uses the decision path scoring function and the decision path optimization function to screen the decision paths. Then, a dynamic pruning strategy is used to eliminate decision paths that do not meet the dynamic pruning conditions after screening. The dynamic pruning conditions include the decision path threshold and the number of executions. The decision path refers to the flight path of the aircraft in the simulation test.

[0034] And a system formation module: used to physically connect the flight control semi-physical simulation test environment with the flight control simulation intelligent body to form a flight control computer simulation test system.

[0035] The present invention proposes a memory storing a program file, wherein the program file is executed to implement program instructions formed by the method for building a flight control computer simulation test system as described above.

[0036] The present invention provides an electronic device, comprising a processor and a memory coupled to each other, wherein:

[0037] The memory is used to store program instructions formed by the method for building the flight control computer simulation test system;

[0038] The processor is configured to execute program instructions stored in the memory.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] 1. The present invention relates to a method for constructing a flight control computer simulation test system. An intelligent agent is introduced and trained using a structured aircraft simulation test knowledge base to form a flight control simulation intelligent agent. A flight control semi-physical simulation test environment is physically connected to the flight control simulation intelligent agent. The present invention utilizes the flight control simulation intelligent agent to perform test design, test execution control, and result analysis for the flight control semi-physical simulation test environment, thereby achieving the goal of autonomously completing simulation tests without relying on human effort. This avoids human error introduced during flight control computer simulation testing and improves the reliability of simulation test results.

[0041] 2. In the present invention, the decision layer of the intelligent agent uses the scoring function of the decision path and the optimization function of the decision path to screen the decision path, and then uses the dynamic pruning strategy to eliminate the decision paths that do not meet the dynamic pruning conditions after screening; a two-stage optimization mechanism is formed through screening and dynamic pruning strategies to ensure that the retained decision paths have higher execution efficiency, stability and task completion, thereby enabling the intelligent agent to have better planning and reasoning capabilities in simulation testing, and improving the intelligence level and decision response quality of the simulation testing process.

[0042] 3. The method for building a flight control computer simulation test system of the present invention also includes configuring a standard API calling interface, and physically connecting the flight control semi-physical simulation test environment with the flight control simulation intelligent body through the standard API calling interface, so that the information of the flight control semi-physical simulation test environment can be operated and perceived by the flight control simulation intelligent body. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A schematic diagram of a method for constructing a flight control computer simulation test system according to the present invention;

[0044] Figure 2 A schematic diagram of a system for building a flight control computer simulation test system according to the present invention;

[0045] Figure 3 Schematic diagram of the training process of the intelligent agent. DETAILED DESCRIPTION

[0046] The technical solution of the present invention will be further explained below with reference to the accompanying drawings and embodiments, but the present invention is not limited to the embodiments described below.

[0047] Flight control computer simulation testing essentially involves designing a large number of tests for different flight control computer components to verify their design, functionality, algorithmic logic, strategies, and performance. Therefore, the test design, execution, and results analysis processes are universal. Based on the underlying logic, test flow, and dependent data analysis methods of flight control computer simulation testing, this paper proposes a method, system, memory, and electronic equipment for building a flight control computer simulation test system.

[0048] See also Figure 1 The present invention proposes a method for building a flight control computer simulation test system, comprising:

[0049] S1: Construct a structured simulation test knowledge base for aircraft; the structured simulation test knowledge base for aircraft is formed based on the simulation test knowledge base of aircraft. The simulation test knowledge base of aircraft includes the initial scenario information of the flight control computer, the simulation test purpose information of the flight control computer, the simulation test method information of the flight control computer, the simulation test process information of the flight control computer, the simulation test data analysis method of the flight control computer, the simulation test result judgment standard of the flight control computer, the sensor data of the flight control computer, the simulation case analysis of the flight control computer, the simulation test outline of the flight control computer, the simulation test system user manual of the flight control computer, and the simulation test report of the flight control computer and other simulation test information.

[0050] Preferably, the present invention also includes building a data processing framework, which includes a data sorting algorithm, a data cleaning strategy, a data segmentation algorithm and a standardization processing algorithm. The data processing framework is used to classify and sort the aircraft's simulation test knowledge base, clean noise, segment data and label data in turn, thereby converting the simulation test information into JSON structured data. The JSON structured data is used as the aircraft's structured simulation test knowledge base, and the aircraft's structured simulation test knowledge base is used to train the intelligent agent, making it easier for the intelligent agent to understand and use it.

[0051] S2: Build a flight control semi-physical simulation test environment and an intelligent agent that operates and perceives the flight control semi-physical simulation test environment. The intelligent agent is trained through the structured simulation test knowledge base of the aircraft.

[0052] In the present invention, the language model structure of the intelligent agent can be selected from one of the open source large models, such as Tongyi Qianwen, iFlytek Spark, Wenxin Yiyan, Zhipu Qingyan and Tencent Hunyuan. Preferably, the language model structure of the intelligent agent in the present invention is Wenxin Yiyan 4.5. The present invention selects the Transformer framework in Wenxin Yiyan 4.5 to train through the structured simulation test knowledge base of the aircraft, so that the Transformer framework predicts the next word in the sentence or fills the word in the vacancy in the sentence, etc. The specific training process is as follows: Figure 3 As shown in the figure, the process includes building a structured simulation test knowledge base of the aircraft, inputting questions, starting training, similarity matching, relevance extraction, word segmentation and completion, and training evaluation, thereby forming a flight simulation intelligent agent.

[0053] In the present invention, the flight control semi-physical simulation test environment is composed of a flight control computer, a real-time simulator, a dynamic model, a sensor model, simulation management software, simulation scheduling software and various buses. Specifically, building a flight control semi-physical simulation test environment is a technology well known to those skilled in the art and will not be repeated in the present invention.

[0054] S3: Build a perception layer, decision layer, memory layer, and execution layer for the intelligent agent to form a flight control simulation intelligent agent. Among them, the decision layer of the intelligent agent uses the scoring function and the optimization function of the decision path to screen the decision path, and then uses the dynamic pruning strategy to eliminate the decision paths that do not meet the dynamic pruning conditions after screening. The dynamic pruning conditions include the decision path threshold and the number of executions. The decision path refers to the flight path of the aircraft in the simulation test.

[0055] S4: Physically connect the flight control semi-physical simulation test environment with the flight control simulation intelligent agent to form a flight control computer simulation test system.

[0056] The method for building the flight control computer simulation test system of the present invention also includes configuring a standard API calling interface, and the flight control semi-physical simulation test environment and the flight control simulation intelligent body are physically connected through the standard API calling interface. The standard API calling interface is oriented to the flight control simulation intelligent body, and it is a tool for the flight control simulation intelligent body to operate and perceive the flight control semi-physical simulation test environment.

[0057] The specific process of configuring the standard API calling interface is: unitize the various control logics in the flight control semi-physical simulation test environment to generate control units, each of which can describe a single function; standardize each control unit to form a standard API calling interface, ensure that each standard API calling interface can be registered by the flight control simulation intelligent agent, and the flight control simulation intelligent agent does not need to do too much thinking and learning when calling the standard API calling interface. Among them, the standard API calling interface includes simulation scene initialization API, simulation scheduling API, fault injection API, model operation API and equipment control API; standardize the data query structure of the flight control semi-physical simulation test environment, convert natural language data into structured data, and form a knowledge base query interface and a data query interface. The structure of the structured data is the same as the structure of the intelligent agent parsed data, ensuring that the queried data can be understood efficiently.

[0058] In the present invention, a perception layer is built for the intelligent agent to enable it to have data acquisition capabilities, specifically including: summarizing the knowledge base query interface and the data query interface, establishing a data dictionary, and associating the perception layer and the decision layer of the intelligent agent through the data dictionary, so that the decision layer of the intelligent agent can call the information perceived in the perception layer of the intelligent agent. The perception layer of the intelligent agent adopts a data search algorithm. The data search algorithm of the present invention is preferably a depth-first algorithm, which matches the decision layer of the intelligent agent, and can ensure that the logic of the data search is consistent with the logic of the decision layer of the intelligent agent, which is a chain logic.

[0059] In the present invention, building a decision layer for the intelligent agent includes using the scoring function of the decision path and the optimization function of the decision path to screen the decision path, and then using the dynamic pruning strategy to eliminate the decision paths that do not meet the dynamic pruning conditions after screening. The dynamic pruning conditions include the decision path threshold and the number of executions.

[0060] The screening using the scoring function of the decision path and the optimization function of the decision path specifically includes:

[0061] The total score of each decision path is determined based on the node action of each decision path and the reward obtained after the node action is executed;

[0062] Scoring function for decision paths for:

[0063]

[0064] Where, , is the i-th decision path of the aircraft in the simulation test experiment. Each decision path consists of a set of node actions executed sequentially; is the t-th node action on the i-th decision path, t≥1, which corresponds to the specific operation instruction executed by the flight control computer in the simulation test, such as initializing the navigation module, starting flight status acquisition, injecting inertial navigation faults, or switching control laws; T is the total number of node actions on the i-th decision path, that is, the total number of operation steps in the simulation test; is a decision factor used to weight the cumulative reward scores obtained from each node action, representing the importance of long-term actions in the decision path. Its value is usually set in the range of [0,1], mainly considering the stability of the strategy in the aircraft control system; r t is the reward score of the t-th node action, which is directly related to the accuracy and completion of the node action and its positive contribution to the mission goal. For example, when the flight control computer successfully identifies the inertial navigation anomaly and switches to the backup navigation, r t is a positive value (such as +1.0); otherwise, if the execution fails or there is no response, r t is a negative value (such as −1.0); is the penalty weight parameter; in the aircraft control system, the decision path with long execution time, high resource consumption or high task complexity will bear a higher penalty weight; is the optimization function for the i-th decision path of the aircraft in the simulation test, which is used to measure the comprehensive cost required by this path to complete the test task, including computing resource consumption (such as CPU usage), number of sensor reading calls, control response delay, and total execution time of the test scenario;

[0065] Penalty weight parameter The formula used to balance the reward and cost of each decision path is:

[0066]

[0067] Where, is the average reward of historical decision paths; is the average cost of the historical decision path.

[0068] Construct a cost function, normalize the cost function, and form an optimization function for the decision path;

[0069] The cost function is used to judge the cost required to complete a single decision path, including time consumption and resource consumption. The specific calculation formula is:

[0070]

[0071] Where, is the cost function of the i-th decision path of the aircraft in the simulation test; t is the sequence number of the node action on the i-th decision path; T is the total number of node actions on the i-th decision path; is the adjustment coefficient of time consumption, which is generally set to 0.6; is the adjustment coefficient of resource consumption, which is generally set to 0.4; is the time taken for the t-th node action; is the resource consumption of the t-th node action of the aircraft in the simulation test.

[0072] The purpose of normalizing the cost function is to solve the problem of score imbalance caused by inconsistent cost scales of different decision paths. Normalizing to a reasonable range, we obtain the optimization function of the decision path. The specific formula is as follows:

[0073]

[0074] Where, is the optimization function of the i-th decision path of the aircraft in the simulation test; is the cost function of the i-th decision path of the aircraft in the simulation test; is the known minimum cost of the historical decision path; is the known maximum cost of the historical decision path.

[0075] The total scores of all decision paths are judged according to the optimization function of the decision path, and the decision paths are screened to form the screened decision paths.

[0076] In order to avoid the influence of the length of the decision path (i-th decision path) on the score, the scoring function of the decision path is Divide by the total number of node actions T on the i-th decision path to get the average score of the node actions on the i-th decision path , in order to determine whether the decision path is a valid path, the formula is:

[0077] In order to prevent the excessive expansion of the decision layer generated by training and maintain high efficiency, a dynamic pruning strategy is used to eliminate decision paths that do not meet the dynamic pruning conditions after screening. Specifically, it includes: constructing a validity pruning formula for the decision path, dynamically evaluating the screened decision paths according to the validity pruning formula, and eliminating decision paths that do not meet the dynamic pruning conditions in the screened decision paths based on the dynamic evaluation results.

[0078] The validity pruning formula is:

[0079] and

[0080] Where, The scoring threshold of the decision path is used to set the lower limit of the score for determining whether the decision path is effective, reflecting the tolerance requirements of the task (node ​​action) in the simulation test for the stability and effectiveness of the decision path. In the simulation test, if the average score of a decision path is lower than this scoring threshold, it will be considered as a poor decision path and will be cut off to avoid interference or misleading the aircraft control system; S i The total number of times the aircraft successfully executes node actions in the i-th decision path in the simulation test, that is, the number of times the decision path completes the predetermined node actions and outputs acceptable results. Usually, the system state is stable, the flight control responds correctly, or the expected data is successfully output as the judgment criteria; n i is the total number of node actions executed by the aircraft in the ith decision path in the simulation test, including success, failure, interruption and other situations; combined with S i , which can be used to calculate the execution success rate of the decision path; N min is the minimum effective execution number, which is an important parameter for measuring the statistical reliability of the decision path evaluation results. If the execution number of a decision path is less than N min Even if the current score is high, it may be regarded as an unstable decision path due to insufficient samples and thus be eliminated by the dynamic pruning strategy. This value is generally set according to the simulation test cycle and task complexity. For example, setting it to 100 times can ensure that the path has basic statistical significance.

[0081] In this invention, the decision layer of the agent quantitatively evaluates and prioritizes multiple feasible decision paths using a decision path scoring function. A total score is first calculated based on the reward and cost of each decision path's node actions during the simulation test task. The cost is normalized to form a decision path optimization function, which adjusts for evaluation biases between different decision paths caused by inconsistent dimensions, making the scoring mechanism more objective and consistent. The agent's decision layer then filters the selected decision paths based on the total score, eliminating inefficient or invalid decision paths. Furthermore, a dynamic pruning strategy is employed to perform a secondary compression on the selected decision paths. Paths with poor stability and low experimental value are identified based on their known minimum cost, known maximum cost, average reward, and average cost. This two-stage optimization mechanism, formed by screening and dynamic pruning, ensures that the remaining decision paths have high execution efficiency, stability, and task completion. This enables the agent to possess superior planning and reasoning capabilities during simulation testing, improving the intelligence level and decision-making response quality of the simulation testing process.

[0082] In the present invention, a memory layer is built for the intelligent agent, including long-term memory and short-term sessions, specifically including: building a short-term session based on the context management mode to store the initial scene settings, test process, test data and test results of a single simulation test in the flight control computer simulation; using a vector database to build long-term memory to store basic knowledge, decision-making information, training experience, scenario information, skill information and various mapping relationships in the flight control computer simulation.

[0083] In the present invention, an execution layer is built for the intelligent agent to build its ability to perform specific actions, and the standard API calling interface is registered to the intelligent agent and identified so that it can understand the function of the standard API calling interface and call it during the execution of the task; a data analysis algorithm library is established to provide an algorithm basis for the execution layer. The data analysis algorithm library includes simulation data preprocessing algorithms, data statistical analysis algorithms and specific algorithms commonly used in simulation, among which the simulation data preprocessing algorithms include data interception, bad pixel removal and smoothing processing, etc.; the data statistical analysis algorithms include mean, variance, RMS value and test algorithm, etc.; the specific algorithms commonly used in simulation include spectrum analysis, filtering analysis and Fourier transform, etc.

[0084] The following describes the method for constructing the flight control computer simulation test system of the present invention in detail by taking the autonomous completion of the simulation test inertial navigation redundancy test process as an example.

[0085] 1) The user inputs the user requirement "test whether the flight control computer's redundancy judgment logic for inertial navigation is correct" to the flight control simulation agent;

[0086] 2) The flight control simulation agent receives user requirements and interprets them. The specific understanding process is as follows:

[0087] ① The flight control simulation agent extracts the core of the target, including inertial navigation: using accelerometers and gyroscopes to determine the position, velocity, and attitude of an object without relying on external signals; redundancy determination: in the case of redundant sensors, determining whether there is a sensor failure or failure and taking appropriate strategies;

[0088] ② The flight control simulation agent defines the test objectives and verifies whether the redundancy judgment logic in the flight control computer is correct. Possible tests include detection under normal conditions, single-point failures, multiple-point failures, progressive failures, and sensor noise and interference.

[0089] 3) The decision-making layer of the intelligent agent divides the task and designs the experimental steps;

[0090] ① The test steps include: flight control computer scenario initialization, flight control computer simulation scheduling, flight control computer data acquisition, flight control computer fault injection, flight control computer result analysis, and flight control computer report generation, thereby conducting the test design;

[0091] ② Evaluate whether the test steps are executable based on the agent's decision layer. If not, call external tools and update the agent's decision layer to execute the test;

[0092] ③ Summarize the test results and analyze the performance of the redundancy judgment logic in different situations. If there are any problems, optimize and adjust them accordingly. The entire process must ensure that the data is accurate and reliable. It may be necessary to repeat the test multiple times to confirm the stability of the results and then conduct result analysis.

[0093] 4) The flight control simulation agent executes the test steps and maps the test tasks to specific standard API call interfaces through the data dictionary to complete the experimental steps. Specifically:

[0094] ①Call the simulation scene initialization API to complete the simulation initialization task;

[0095] ②Call the simulation scheduling API to control the simulation computer and dynamic model of the flight control semi-physical simulation test environment to start the simulation test;

[0096] ③Call the data query interface to collect the position and attitude data of the main inertial navigation, backup inertial navigation, flight control, etc.;

[0097] ④ Call the fault injection API to introduce errors to the main inertial navigation system, such as increased bias, noise, or complete failure;

[0098] ⑤ Repeat steps ①-⑤ above to combine multiple fault scenarios.

[0099] 5) The flight control simulation agent records various test data from the beginning to the end of the simulation test through short-term memory, including: decision execution status, time consumption, resource consumption and other test data.

[0100] 6) The flight control simulation agent analyzes the test data to determine whether the flight control computer redundancy determination logic is correct, including:

[0101] ① During normal simulation, monitor the main inertial navigation position and attitude information of the flight control computer to determine whether the flight control computer's redundancy judgment logic for inertial navigation is correct under normal circumstances;

[0102] ② After injecting a fault into the primary inertial navigation system of the flight control computer, the flight control computer was monitored and found to be using the backup inertial navigation position and attitude information. This confirmed that the flight control computer's redundancy judgment logic for inertial navigation in the event of a primary inertial navigation failure was correct.

[0103] ③ Similarly, multi-point fault testing, progressive fault testing (signal gradually weakening), noise and interference testing, etc. can be performed to verify whether its judgment logic is correct.

[0104] 7) The flight control simulation agent calls the simulation test report template to generate a test report, which includes:

[0105] ① The execution layer of the intelligent agent obtains the simulation test report template from the database responsible for long-term memory;

[0106] ② The execution layer of the intelligent agent fills the test steps, test scenarios, test results and conclusions and other data into the template to generate a test report.

[0107] The following describes the process of building a decision-making layer for an intelligent agent using specific data:

[0108] According to the above autonomous simulation test inertial navigation redundancy test process, the generated decision path is scored and the decision factor is set. is 0.9, and the reward score r of the t-th node action t See Table 1.

[0109] Table 1: Reward scores for node actions on the decision path

[0110] Execution Results Reward Points Node action execution successful + data complete +1.0 Node action partially successful / data incomplete +0.5 Node action execution failed / no response -1.0

[0111] Combined with the historical data of flight control computer simulation tests, the average reward of historical decision paths is 0.75, the average cost of the historical decision path is 0.6, so the penalty weight parameter is 0.625.

[0112] Combined with the characteristics of flight control computer simulation test, considering that the proportion of time consumption is higher than the proportion of resource consumption, therefore, set ; Record the execution status of the decision path as shown in Table 2.

[0113] Table 2: Implementation of decision paths

[0114] Serial number Execution steps Reward Points Node action time Node action resource consumption 1 Call the model initialization and simulation scene initialization API to complete the simulation initialization task 1 10 21 2 Call the simulation start API to control the simulation computer and dynamic model to start simulation 1 12 17 3 Call the data acquisition interface to collect the position and attitude data of the main inertial navigation, backup inertial navigation, flight control, etc. 0.5 20 25 4 Call the fault injection API to introduce errors to the main inertial navigation system, such as bias, noise increase, or complete failure 1 14 10 5 The agent analyzes the test data to determine whether the flight control computer's redundancy determination logic is correct; 1 20 16 6 The intelligent agent calls the simulation test report template to generate a test report 0.5 40 43

[0115] Substituting the data in Table 2 into the cost function, we obtain that the execution cost of the i-th decision path is 122.4.

[0116] The cost function is normalized to obtain the optimization function of the decision path. The optimized execution cost calculated by the optimization function of the decision path is 0.312.

[0117] Scoring function through decision path The total score of the decision path generated by this simulation experiment is 3.7903.

[0118] The decision path of this simulation experiment is evaluated according to the scoring judgment table. The scoring judgment table is shown in Table 3.

[0119] Table 3: Scoring table

[0120] Average rating Judgment results Strategy 0.4 or less Ineffective decision chain give up 0.4—0.6 Partially effective decision chain The update enters the decision chain and is marked as optimizable, and continues to be evaluated in subsequent trial training. 0.6 Effective decision-making chain Update into the decision chain

[0121] It can be seen from Table 3 that the judgment result of the decision path in this simulation experiment is valid.

[0122] The effectiveness pruning formula is used to perform 200 tests on the decision path of the decision layer of the intelligent agent, and the execution results are recorded as: success: 191 times, failure: 9 times. It is judged that the decision path does not meet the pruning conditions and is not pruned.

[0123] The present invention utilizes a flight control simulation intelligent agent to perform test design, test execution control, and result analysis for a flight control semi-physical simulation test environment, thereby achieving the purpose of autonomously completing simulation tests without relying on manpower, avoiding human errors introduced during the flight control computer simulation test process, and improving the reliability of the simulation test results.

[0124] See also Figure 2 The present invention also provides a flight control computer simulation test system construction system formed based on the above-mentioned flight control computer simulation test system construction method, including a construction module, a first construction module, a second construction module and a system formation module;

[0125] Building module: used to build a structured simulation test knowledge base for aircraft;

[0126] The first building module is used to build a flight control semi-physical simulation test environment and an intelligent agent that operates and perceives the flight control semi-physical simulation test environment. The intelligent agent is trained using a structured simulation test knowledge base for the aircraft.

[0127] The second building module is used to build the perception layer, decision layer, memory layer, and execution layer for the intelligent agent to form a flight control simulation intelligent agent. The decision layer of the intelligent agent uses the decision path scoring function and the decision path optimization function to screen the decision paths. Then, a dynamic pruning strategy is used to eliminate decision paths that do not meet the dynamic pruning conditions after screening. The dynamic pruning conditions include the decision path threshold and the number of executions. The decision path refers to the flight path of the aircraft in the simulation test.

[0128] System formation module: used to physically connect the flight control semi-physical simulation test environment with the flight control simulation intelligent body to form a flight control computer simulation test system.

[0129] The construction system of the flight control computer simulation test system of the present invention corresponds to the construction method of the above-mentioned flight control computer simulation test system. Among them, the specific contents of the construction module, the first construction module, the second construction module and the system formation module refer to the description of the construction method of the above-mentioned flight control computer simulation test system, and the present invention will not repeat them here.

[0130] The present invention also provides a memory storing a program file, wherein the program file is executed to implement the program instructions formed by the method for building the flight control computer simulation test system. Specifically, the method for building the flight control computer simulation test system is described above.

[0131] The memory in the present invention may specifically include random access memory (RAM), internal memory, read-only memory (ROM), programmable ROM, erasable programmable ROM, registers, a hard disk, a removable disk, or a CD-ROM. It should be noted that those skilled in the art may select the form and type of storage medium based on actual use requirements, and the present invention does not further specify this.

[0132] The present invention also provides an electronic device comprising a processor and a memory coupled to each other, wherein the memory is configured to store program instructions generated by the aforementioned method for constructing a flight control computer simulation test system, and the processor is configured to execute the program instructions stored in the memory. For details on the method for constructing a flight control computer simulation test system, please refer to the description above.

[0133] The electronic device in the present invention includes any electronic device that can execute program instructions, such as a computer, a mobile terminal, a remote control device or a wearable device.

[0134] The above content is only used to illustrate the technical solution of the present invention, rather than to limit the present invention. Although the present invention has been described in detail with reference to the foregoing, it should be understood by those skilled in the art that the technical solution described above can still be modified, or some or all of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solution to deviate from the scope of the technical solution of the present invention.

Claims

1. A method for constructing a flight control computer simulation test system, characterized in that: include: Build a knowledge base for structured simulation testing of aircraft; Build a flight control semi-physical simulation test environment and an intelligent agent that operates and perceives the flight control semi-physical simulation test environment. The intelligent agent is trained using a structured simulation test knowledge base for the aircraft. The intelligent agent is constructed with a perception layer, a decision layer, a memory layer, and an execution layer to form a flight control simulation intelligent agent. The decision layer of the intelligent agent uses a scoring function and an optimization function to screen decision paths. Then, a dynamic pruning strategy is used to remove decision paths that do not meet the dynamic pruning conditions after screening. The dynamic pruning conditions include the decision path threshold and the number of executions. The decision path refers to the flight path of the aircraft in the simulation test. The flight control semi-physical simulation test environment is physically connected to the flight control simulation intelligent body to form a flight control computer simulation test system.

2. The method for constructing a flight control computer simulation test system according to claim 1, characterized in that: The screening using the scoring function of the decision path and the optimization function of the decision path specifically includes: The total score of each decision path is determined based on the node action of each decision path and the reward obtained after the node action is executed; Construct a cost function, normalize the cost function, and form an optimization function for the decision path; The total scores of all decision paths are judged according to the optimization function of the decision path, and the decision paths are screened to form the screened decision paths.

3. The method for constructing a flight control computer simulation test system according to claim 2, characterized in that: The decision paths that do not meet the dynamic pruning conditions after being screened by using the dynamic pruning strategy specifically include: A validity pruning formula for decision paths is constructed, and the screened decision paths are dynamically evaluated according to the validity pruning formula. Decision paths that do not meet the dynamic pruning conditions are eliminated from the screened decision paths based on the dynamic evaluation results.

4. The method for constructing a flight control computer simulation test system according to claim 3, wherein: The construction of a structured simulation test knowledge base for an aircraft specifically includes: Obtain the aircraft simulation test knowledge base; The data organization algorithm, data cleaning strategy, data segmentation algorithm and standardization processing algorithm are used to classify, organize, clean noise, segment data and label data in sequence to form a structured simulation test knowledge base of aircraft.

5. The method for constructing a flight control computer simulation test system according to claim 1, wherein: The method for constructing the flight control computer simulation test system also includes establishing a data dictionary, and associating the perception layer and decision layer of the intelligent body through the data dictionary, so that the decision layer of the intelligent body can call the information perceived by the perception layer of the intelligent body.

6. The method for constructing a flight control computer simulation test system according to claim 1, characterized in that: The method for building the flight control computer simulation test system also includes configuring a standard API calling interface, and the flight control semi-physical simulation test environment and the flight control simulation intelligent body are physically connected through the standard API calling interface.

7. The method for constructing a flight control computer simulation test system according to claim 4, characterized in that: The simulation test knowledge base of the aircraft includes initial scenario information of the flight control computer, simulation test purpose information of the flight control computer, simulation test method information of the flight control computer, simulation test process information of the flight control computer, simulation test data analysis method of the flight control computer, simulation test result judgment criteria of the flight control computer, sensor data of the flight control computer, simulation case analysis of the flight control computer, simulation test outline of the flight control computer and simulation test report of the flight control computer.

8. A flight control computer simulation test system construction system formed based on the flight control computer simulation test system construction method according to claim 1, characterized in that: include: Building module: used to build a structured simulation test knowledge base for aircraft; The first building module is used to build a flight control semi-physical simulation test environment and an intelligent agent that operates and perceives the flight control semi-physical simulation test environment. The intelligent agent is trained using a structured simulation test knowledge base for the aircraft. The second building module is used to build the perception layer, decision layer, memory layer, and execution layer for the intelligent agent to form a flight control simulation intelligent agent. The decision layer of the intelligent agent uses the decision path scoring function and the decision path optimization function to screen the decision paths. Then, a dynamic pruning strategy is used to eliminate decision paths that do not meet the dynamic pruning conditions after screening. The dynamic pruning conditions include the decision path threshold and the number of executions. The decision path refers to the flight path of the aircraft in the simulation test. And a system formation module: used to physically connect the flight control semi-physical simulation test environment with the flight control simulation intelligent body to form a flight control computer simulation test system.

9. A memory, characterized in that: A program file is stored, and the program file is executed to implement program instructions formed by the method for building a flight control computer simulation test system according to any one of claims 1 to 7.

10. An electronic device, characterized in that: The system comprises a processor and a memory coupled to each other, wherein: The memory is used to store program instructions formed by the method for building a flight control computer simulation test system according to any one of claims 1 to 7; The processor is configured to execute program instructions stored in the memory.

Images