Overall distributed control method of diesel engine high gas environment test bench

Abstract

This invention discloses a distributed control method for a high-gas environment test bench for diesel engines, relating to the field of detection and control. The method includes the following steps: Step 1: Define the controlled components within the diesel engine to be tested, including operating and safety components, that require monitoring. Deploy corresponding sensing devices based on the characteristics of these components to collect core performance parameters of each controlled component during the test in a distributed manner. Step 2: According to the test requirements, set target gas concentration, target temperature, and target pressure for the test bench chamber. Based on these set values, perform gas, temperature, and pressure distribution within the chamber. This method can identify early abnormal trends, issue differentiated adjustment commands to different controlled components, and form a distributed collaborative response. Based on the collaborative control logic of global situation assessment, it effectively avoids malfunctions or insufficient response, ensuring dynamic optimization and adaptive adjustment of the control strategy.

Description

Technical Field

[0001] This invention relates to the field of detection and control technology, specifically to an overall distributed control method for a high-gas environment test bench for diesel engines. Background Technology

[0002] In underground engineering fields such as coal mines and tunnels, the safety requirements for power equipment are extremely stringent. In these environments, diesel engines are an important power source, but their operation faces the inherent and deadly risk of methane buildup. Methane, as a flammable and explosive gas, can ignite upon contact with an open flame or high-temperature surface within a certain concentration range, posing a significant threat to life and property. Therefore, before deploying diesel engines in such environments, highly simulated tests must be conducted to verify their operational safety, reliability, and the effectiveness of associated safety devices under preset high methane concentrations.

[0003] Existing control methods for high-gas environment testing of diesel engines often have the following shortcomings: Traditional methods often focus on controlling environmental parameters such as gas concentration and temperature, or only independently monitor a few key points of the diesel engine. They are unable to integrate environmental simulation, main engine operating status, and safety component response as an organic whole for coordinated analysis and control. They cannot identify potential risks indicated by the dynamic evolution of multiple parameters, have insufficient early warning capabilities for abnormal states, and rely heavily on the experience of operators for final safety judgments. They lack intelligent comprehensive risk quantification and prediction capabilities, making it difficult to integrate and evaluate the operational health and safety system effectiveness of the diesel engine under complex operating conditions. This results in passive and lagging control strategies, failing to achieve adaptive and proactive adjustments based on real-time risk assessment, and appearing insufficiently sensitive and reliable in dealing with sudden or gradual risks. Summary of the Invention

[0004] (a) Technical problems to be solved In view of the above-mentioned shortcomings of the existing technology, the present invention provides an overall distributed control method for a high-gas environment test bench for diesel engines, which can effectively solve the problems of the existing technology.

[0005] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: This invention discloses a distributed control method for a high-gas environment test bench for diesel engines, comprising the following steps: Step 1: Define the controlled components within the diesel engine to be tested, including operating and safety components, that need to be monitored. Deploy corresponding sensing devices based on the characteristics of these components to collect the core performance parameters of each controlled component during the test in a distributed manner. Step 2: According to the test requirements, set the target gas concentration, target temperature, and target pressure for the test bench; control the gas distribution system, temperature regulation system, and pressure regulation system to work together to distribute gas, temperature, and pressure into the test bench according to the set values ​​to simulate a high-gas environment; the gas distribution system adopts a dynamic proportional mixing method, and uses high-precision flow meters and concentration sensors to achieve precise and stable control of the gas concentration in the test bench; the temperature and pressure distribution processes are carried out synchronously with the gas distribution process to simulate the comprehensive environmental conditions in a real mine. Step 3: In the current simulated high-gas environment, start the diesel engine to be tested according to the preset operating settings; use sensing devices to obtain the surface temperature or exhaust temperature of each controlled component in real time, identify and record the temperature change curves over time; Step 4: Extract the temperature fluctuation characteristics that characterize the operating status from the acquired temperature change curve; when the set target gas concentration exceeds the preset safety threshold, synchronously collect the response time of the safety component from receiving the over-limit signal to completing the preset safety action, and evaluate its response effect coefficient based on the execution effect of the safety action; Step 5: Construct a security identification model based on deep learning algorithms and train the model using historical test data. Use the temperature fluctuation characteristics of several key authorized components, the response time of security components, and the response effect coefficients extracted during the current test cycle as input to the model. Step 6: The safety identification model performs fusion analysis and risk assessment on the input parameters and outputs a comprehensive safety coefficient. This safety coefficient is used to quantitatively characterize the overall safety level of the diesel engine and related controlled components under the current test conditions. Step 7: Using the safety factor as the core feedback parameter, compare it with the preset safety threshold of each associated controlled component. Based on the comparison results, generate distributed adjustment instructions for different controlled components to adaptively adjust the operating parameters of the diesel engine or the response logic of the safety components, so as to achieve safety control of the test process.

[0006] Furthermore, the operating components in step 1 include the cylinder head, piston, fuel injector, turbocharger, and exhaust pipe; the safety components include an automatic flameout device for excessive gas concentration, an intake flame arrester, and an emergency pressure relief valve; and the core performance parameters include temperature, vibration, pressure, and speed.

[0007] Furthermore, the extraction process of temperature fluctuation features in step 4 includes the following steps: The raw temperature change curve data collected in real time by the sensing device is filtered, denoised, and outlier removed to obtain a smooth temperature change curve. Based on the diesel engine's speed and load operating parameters, the test process was divided into multiple steady-state and transient operating condition intervals. For different operating conditions, feature parameters are extracted from the time domain, frequency domain, and time-frequency domain respectively. In the time domain, the rate of rise, average slope, fluctuation variance, and duration exceeding the preset alarm threshold of the temperature change curve per unit time are extracted. In the frequency domain, a fast Fourier transform is performed on the temperature signal to extract the energy proportion or main frequency amplitude within a specific frequency band. In the time-frequency domain, wavelet transform is used to extract the energy distribution characteristics of the temperature signal at different scales. The multi-dimensional feature parameters extracted from the time domain, frequency domain, and time and frequency domain are normalized, and principal component analysis is used to fuse features and reduce dimensionality to form a core temperature fluctuation feature vector that can characterize the current operating state.

[0008] Furthermore, the response time acquisition process in step 4 is as follows: A command listening node is set up in the control loop of the safety component. When the distributed controller issues a safety action command, a first time stamp is generated and written to the log. A motion feedback sensor is deployed at the execution terminal of the safety component. When the sensor detects that the component displacement has reached a preset termination point or the state switch is completed, a second time stamp is generated. The time difference between the first and second time scales is calculated as the response time, where: For valve-type safety components, a high-precision displacement sensor is used to detect the movement trajectory of the valve core, with the valve core contacting the valve seat sealing surface as the termination point. For circuit protection safety components, the signal for contact switching is detected by a relay status sensor.

[0009] Furthermore, the evaluation process of the response effect coefficient in step 4 includes the following steps: Obtain a quantitative value of the completion of the safety component's action. This quantitative value is determined by measuring the ratio of the actual displacement to the rated stroke, or by converting the deviation between the pressure feedback signal at the execution endpoint and the target pressure. Calculate the response time utility value, which decays exponentially with increasing response time, and the decay rate is controlled by a pre-configured decay factor; The quantified value of action completion and the utility value of response time are added together with preset weights, where the weight coefficient of action completion is greater than the weight coefficient of response time. Each weight coefficient is determined through calibration experiments.

[0010] Furthermore, the construction process of the security identification model in step 5 includes the following steps: Step 51: Collect historical test data; Step 52: Preprocess the temperature time series in the historical experimental data and extract the temperature fluctuation features. Combine these features with the corresponding response time and response effect coefficient to form a multi-dimensional feature vector, which serves as the input feature set for the model. Step 53: Select a deep learning network containing a time series processing module and a fully connected analysis module as the model architecture; divide the input feature set into a training set and a validation set according to the ratio, and use the training set to train the model; Step 54: Use the validation set to evaluate the accuracy and recall performance metrics of the trained model. Once the preset performance requirements are met, solidify the model into an executable security identification model and deploy it to the control system of the test bench for online real-time calculation of the security factor.

[0011] Furthermore, the historical data includes the temperature time series of each authorized component, the response time and effect evaluation results of the safety components, and the manual annotation of the safety level of the corresponding test results under different gas concentrations, temperatures, and pressures.

[0012] Furthermore, during the model training process, the training set iteratively adjusts the model parameters through an optimization algorithm to minimize the loss function between the predicted safety level output by the model and the manually labeled safety level.

[0013] Furthermore, the formula for calculating the safety factor in step 6 is as follows: ; In the formula, This represents the safety factor of the final output, with a value range of [0, 1]. Represents the characteristic vector of temperature fluctuation. Represents the temperature fluctuation feature vector of the input. After standardization and normalization, the temperature state safety score calculated by the sub-network in the safety identification model has a value range of [0,1]. This represents a decay function based on response time, indicating that the longer the response time, the lower the contribution of this term. Represents the response time of the safety component, in seconds. The response performance coefficient representing the safety component is a normalized value obtained based on the performance evaluation, and its value ranges from [0,1]. This represents the response time decay factor, a constant greater than 0, used to adjust the sensitivity of the response time to the safety factor. Weighting coefficients representing temperature state. The weighting coefficient representing response time. Weighting coefficients representing the response effect.

[0014] Furthermore, the distributed adjustment instructions in step 7 include: When the safety factor is lower than the first threshold, a command is sent to the diesel engine electronic control unit to limit its output power or speed by a preset range. When the safety factor falls below a lower second threshold, a pre-action command is sent directly to the safety component or a higher-level emergency protection procedure is triggered, while all current parameters are recorded for failure analysis.

[0015] (III) Beneficial Effects Compared with the known prior art, the technical solution provided by this invention has the following beneficial effects: 1. By systematically deploying a distributed sensor network, not only is the temperature dynamics of key operating components of the diesel engine collected in real time, but the response performance of safety components is also monitored simultaneously. This constructs a multi-dimensional data acquisition system. By introducing a safety identification model based on deep learning, discrete temperature fluctuation characteristics, precise response time, and effect coefficients are deeply integrated and analyzed to output a quantitative safety coefficient. This model can identify early abnormal trends and provide accurate early warnings before potential risks evolve into actual failures or dangers.

[0016] 2. By using the safety factor as the core intelligent criterion, the system achieves full-process linkage control from environmental simulation and diesel engine operation to safety response. Based on the unified safety assessment results, the system can issue differentiated adjustment commands to different controlled components, forming a distributed collaborative response. In the early stage of risk, it actively limits power operation to suppress temperature rise, and in the case of risk aggravation, it orderly triggers multi-level safety protection. Based on the collaborative control logic of global situation assessment, it effectively avoids malfunctions or insufficient response, and ensures the dynamic optimization and adaptive adjustment of the control strategy. Attached Figure Description

[0017] 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.

[0018] Figure 1 This is a schematic diagram of the overall process of the present invention; Figure 2 This is a flowchart illustrating the process of constructing the security identification model in this invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0020] The present invention will be further described below with reference to embodiments.

[0021] The overall distributed control method of the diesel engine high-gas environment test bench in this embodiment is as follows: Figure 1 As shown, it includes the following steps: Step 1: Define the controlled components within the diesel engine to be tested, including operating and safety components, that require monitoring. Deploy corresponding sensing devices based on the characteristics of each component to collect the core performance parameters of each controlled component during the test in a distributed manner. Operating components include the cylinder head, piston, injector, turbocharger, and exhaust pipe. Safety components include an automatic flameout device for excessive gas concentration, an intake flame arrester, and an emergency pressure relief valve. Core performance parameters include temperature, vibration, pressure, and speed.

[0022] Step 2: According to the test requirements, set the target gas concentration, target temperature, and target pressure for the test bench; control the gas distribution system, temperature regulation system, and pressure regulation system to work together to distribute gas, temperature, and pressure in the test bench according to the set values ​​to simulate a high-gas environment; the gas distribution system adopts a dynamic proportional mixing method, and uses high-precision flow meters and concentration sensors to achieve precise and stable control of the gas concentration in the test bench; the temperature and pressure distribution processes are carried out synchronously with the gas distribution process to simulate the comprehensive environmental conditions in a real mine.

[0023] Step 3: In the current simulated high-gas environment, start the diesel engine to be tested according to the preset operating settings; use sensing devices to obtain the surface temperature or exhaust temperature of each controlled component in real time, identify and record the temperature change curves over time.

[0024] Step 4: Extract temperature fluctuation characteristics representing the operating status from the acquired temperature change curve; when the set target gas concentration exceeds the preset safety threshold, simultaneously collect the response time of the safety component from receiving the over-limit signal to completing the preset safety action, and evaluate its response effect coefficient based on the execution effect of the safety action; the temperature fluctuation characteristic extraction process includes the following steps: The raw temperature change curve data collected in real time by the sensing device is filtered, denoised, and outlier removed to obtain a smooth temperature change curve. Based on the diesel engine's speed and load operating parameters, the test process was divided into multiple steady-state and transient operating condition intervals. For different operating conditions, feature parameters are extracted from the time domain, frequency domain, and time-frequency domain respectively. In the time domain, the rate of rise, average slope, fluctuation variance, and duration exceeding the preset alarm threshold of the temperature change curve per unit time are extracted. In the frequency domain, a fast Fourier transform is performed on the temperature signal to extract the energy proportion or main frequency amplitude within a specific frequency band. In the time-frequency domain, wavelet transform is used to extract the energy distribution characteristics of the temperature signal at different scales. The multi-dimensional feature parameters extracted from the time domain, frequency domain, and time and frequency domain are normalized, and principal component analysis is used to fuse features and reduce dimensionality to form a core temperature fluctuation feature vector that can characterize the current operating state. The process of collecting response time data is as follows: A command listening node is set up in the control loop of the safety component. When the distributed controller issues a safety action command, a first time stamp is generated and written to the log. A motion feedback sensor is deployed at the execution terminal of the safety component. When the sensor detects that the component displacement has reached a preset termination point or the state switch is completed, a second time stamp is generated. The time difference between the first and second time scales is calculated as the response time, where: For valve-type safety components, a high-precision displacement sensor is used to detect the movement trajectory of the valve core, with the valve core contacting the valve seat sealing surface as the termination point. For circuit protection safety components, the signal for contact switching is detected by a relay status sensor.

[0025] Step 5: Construct a security identification model based on deep learning algorithms and train the model using historical test data. Use the temperature fluctuation characteristics of several key authorized components, the response time of security components, and the response effect coefficients extracted during the current test cycle as input to the model.

[0026] Step 6: The safety identification model performs fusion analysis and risk assessment on the input parameters and outputs a comprehensive safety coefficient. This safety coefficient is used to quantitatively characterize the overall safety level of the diesel engine and related controlled components under the current test conditions.

[0027] Step 7: Using the safety factor as the core feedback parameter, compare it with the preset safety threshold of each associated controlled component. Based on the comparison results, generate distributed adjustment instructions for different controlled components to adaptively adjust the operating parameters of the diesel engine or the response logic of the safety components, so as to achieve safety control of the test process. Distributed adjustment instructions include: When the safety factor is lower than the first threshold, a command is sent to the diesel engine electronic control unit to limit its output power or speed by a preset range. When the safety factor falls below a lower second threshold, a pre-action command is sent directly to the safety component or a higher-level emergency protection procedure is triggered, while all current parameters are recorded for failure analysis.

[0028] Compared with existing technologies, this invention deeply integrates high-gas environment simulation, real-time monitoring of multi-dimensional operating parameters, safety component performance evaluation, and deep learning-based intelligent diagnosis of safety risks. It constructs an adaptive control system with a dynamic safety factor as a unified quantitative indicator. This system can comprehensively assess environmental hazards, diesel engine operating status, and safety system reliability, and accordingly perform distributed and precise control of various controlled components. This improves the safety of the test process and the comprehensiveness of the evaluation results, providing an efficient testing platform for the applicability certification and safety boundary exploration of diesel engines in extreme flammable environments.

[0029] At other levels, this embodiment also provides a process for constructing a security identification model, such as... Figure 2 As shown, it includes the following steps: Step 51: Collect historical test data; historical data includes temperature time series of each authorized component, response time and effect evaluation results of safety components under different gas concentrations, temperatures and pressures, and manual annotation of the safety level of the corresponding test results.

[0030] Step 52: Preprocess the temperature time series in the historical experimental data and extract the temperature fluctuation features. Combine these features with the corresponding response time and response effect coefficient to form a multi-dimensional feature vector, which serves as the input feature set for the model.

[0031] Step 53: Select a deep learning network containing a time series processing module and a fully connected analysis module as the model architecture; divide the input feature set into a training set and a validation set according to the proportion, and use the training set to train the model; during the training process of the model on the training set, iteratively adjust the model parameters through optimization algorithms to minimize the loss function between the predicted safety level output by the model and the manually labeled safety level.

[0032] Step 54: Use the validation set to evaluate the accuracy and recall performance metrics of the trained model. Once the preset performance requirements are met, solidify the model into an executable security identification model and deploy it to the control system of the test bench for online real-time calculation of the security factor.

[0033] The formula for calculating the safety factor is: ; In the formula, This represents the safety factor of the final output, with a value range of [0, 1]. Represents the characteristic vector of temperature fluctuation. Represents the temperature fluctuation feature vector of the input. After standardization and normalization, the temperature state safety score calculated by the sub-network in the safety identification model has a value range of [0,1]. This represents a decay function based on response time, indicating that the longer the response time, the lower the contribution of this term. Represents the response time of the safety component, in seconds. The response performance coefficient representing the safety component is a normalized value obtained based on the performance evaluation, and its value ranges from [0,1]. This represents the response time decay factor, a constant greater than 0, used to adjust the sensitivity of the response time to the safety factor. Weighting coefficients representing temperature state. The weighting coefficient representing response time. Weighting coefficients representing the response effect; The operating status of the diesel engine, the response speed and effectiveness of the safety system are quantitatively integrated through a model that combines linear weighting and nonlinear decay. This enables a dynamic and comprehensive assessment of the overall safety of the diesel engine system under high gas conditions. It overcomes the limitations of traditional methods that rely solely on single threshold alarms or isolated parameter judgments, and improves the accuracy and foresight of risk warnings.

[0034] This embodiment provides a process for evaluating the response effect coefficient, including the following steps: Obtain a quantitative value of the completion of the safety component's action. This quantitative value is determined by measuring the ratio of the actual displacement to the rated stroke, or by converting the deviation between the pressure feedback signal at the execution endpoint and the target pressure. Calculate the response time utility value, which decays exponentially with increasing response time, and the decay rate is controlled by a pre-configured decay factor; The quantified value of action completion and the utility value of response time are added together with preset weights, where the weight coefficient of action completion is greater than the weight coefficient of response time. Each weight coefficient is determined by calibration test. For different types of safety components, specific evaluation dimensions are defined and quantified based on their design functions. For automatic flameout devices, the evaluation dimensions include the thoroughness of the flameout action and the reliability of preventing reignition. Thoroughness is assessed by monitoring whether there is any uncut fuel supply or abnormal ignition signal within a specified time window after flameout, and the reliability of preventing reignition is assessed by monitoring the continuous decrease in cylinder and exhaust temperatures and changes in concentration after flameout. For intake flame arresters, the evaluation dimensions include flame-arresting efficiency and additional impact on the intake system. This is assessed by comparing the performance of the flame arrester before and after simulated sparks or detonation. The effectiveness of the flame arrestor is indirectly assessed by observing its condition during operation, and the additional impact is assessed by monitoring the changes in pressure loss in the intake pipe before and after operation. For the emergency pressure relief valve, the evaluation dimensions of its performance include the adequacy of pressure release and the recovery status of the system after the action. Adequacy is assessed by monitoring the time and rate required for the pressure in the control compartment to drop to a safe range after the pressure relief action is triggered, and the system recovery status is assessed by checking whether the seal is intact after the valve is closed. The quantitative monitoring results of each evaluation dimension are combined with the preset weighting relationship and scoring criteria to comprehensively calculate a response effect coefficient value that characterizes the overall performance.

[0035] In summary, this invention, by systematically defining controlled components and deploying sensing devices, achieves distributed parameter acquisition of key operating and safety components of a diesel engine, and constructs an intelligent control architecture that integrates environmental simulation, real-time monitoring and control. It not only monitors routine operating parameters, but also actively collects and quantifies the key performance indicators of the safety protection system in simulated high-risk gas concentration environments, thereby achieving a dual-track parallel assessment of operational risks and safety assurance capabilities. By introducing a safety identification model based on deep learning, it is possible to perform deep fusion and high-order correlation analysis on time-series temperature characteristics, response time, and effect coefficients, and finally output a comprehensive quantitative safety coefficient. This coefficient condenses the complex system safety status into an intuitive indicator and serves as a global feedback signal to drive distributed adjustments to different controlled components, ensuring the safety of the test process and improving the accuracy of the assessment and the level of intelligent control.

[0036] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A distributed control method for a high-gas environment test bench for diesel engines, characterized in that, Includes the following steps: Step 1: Define the controlled components within the diesel engine to be tested, including operating and safety components, that need to be monitored. Deploy corresponding sensing devices based on the characteristics of these components to collect the core performance parameters of each controlled component during the test in a distributed manner. Step 2: According to the test requirements, set the target gas concentration, target temperature and target pressure for the test bench chamber, and perform gas distribution, temperature distribution and pressure distribution in the test bench chamber according to the set values; Step 3: Start the diesel engine to be tested according to the preset operating settings; use sensing devices to obtain the surface temperature or exhaust temperature of each controlled component in real time, identify and record the temperature change curve over time; Step 4: Extract temperature fluctuation characteristics that characterize the operating status from the acquired temperature change curves; When the set target gas concentration exceeds the preset safety threshold, the response time of the safety component from receiving the over-limit signal to completing the preset safety action is collected synchronously, and its response effect coefficient is evaluated based on the execution effect of the safety action. Step 5: Construct a security identification model based on deep learning algorithms, using the extracted temperature fluctuation characteristics of several key authorized components, the response time of security components, and the response effect coefficients as model inputs; Step 6: The security identification model performs fusion analysis and risk assessment on the input parameters and outputs a comprehensive security coefficient; Step 7: Use the safety factor as the core feedback parameter and compare it with the preset safety threshold of each associated controlled component. Based on the comparison results, generate distributed adjustment instructions for different controlled components.

2. The overall distributed control method for the high-gas environment test bench for diesel engines according to claim 1, characterized in that, The operating components in step 1 include the cylinder head, piston, fuel injector, turbocharger, and exhaust pipe; the safety components include an automatic flameout device for excessive gas concentration, an intake flame arrester, and an emergency pressure relief valve; the core performance parameters include temperature, vibration, pressure, and speed.

3. The overall distributed control method for the high-gas environment test bench for diesel engines according to claim 1, characterized in that, The extraction process of temperature fluctuation features in step 4 includes the following steps: The raw temperature change curve data collected in real time by the sensing device is filtered, denoised, and outlier removed to obtain a smooth temperature change curve. Based on the diesel engine's speed and load operating parameters, the test process was divided into multiple steady-state and transient operating condition intervals. For different operating conditions, feature parameters are extracted from the time domain, frequency domain, and time-frequency domain respectively. In the time domain, the rate of rise, average slope, fluctuation variance, and duration exceeding the preset alarm threshold of the temperature change curve per unit time are extracted. In the frequency domain, a fast Fourier transform is performed on the temperature signal to extract the energy proportion or main frequency amplitude within a specific frequency band. In the time-frequency domain, wavelet transform is used to extract the energy distribution characteristics of the temperature signal at different scales. The multi-dimensional feature parameters extracted from the time domain, frequency domain, and time and frequency domain are normalized, and principal component analysis is used to fuse features and reduce dimensionality to form a core temperature fluctuation feature vector that can characterize the current operating state.

4. The overall distributed control method for the high-gas environment test bench for diesel engines according to claim 1, characterized in that, The process of acquiring the response time in step 4 is as follows: A command listening node is set up in the control loop of the safety component. When the distributed controller issues a safety action command, a first time stamp is generated and written to the log. A motion feedback sensor is deployed at the execution terminal of the safety component. When the sensor detects that the component displacement has reached a preset termination point or the state switch is completed, a second time stamp is generated. The time difference between the first and second time scales is calculated as the response time, where: For valve-type safety components, a high-precision displacement sensor is used to detect the movement trajectory of the valve core, with the valve core contacting the valve seat sealing surface as the termination point. For circuit protection safety components, the signal for contact switching is detected by a relay status sensor.

5. The overall distributed control method for the high-gas environment test bench for diesel engines according to claim 1, characterized in that, The evaluation process of the response effect coefficient in step 4 includes the following steps: Obtain a quantitative value of the completion of the safety component's action. This quantitative value is determined by measuring the ratio of the actual displacement to the rated stroke. Calculate the response time utility value, which decays exponentially with increasing response time, and the decay rate is controlled by a pre-configured decay factor; The quantified value of action completion and the utility value of response time are added together with a preset weight, where the weight coefficient of action completion is greater than the weight coefficient of response time.

6. The overall distributed control method for the high-gas environment test bench for diesel engines according to claim 1, characterized in that, The process of constructing the security identification model in step 5 includes the following steps: Step 51: Collect historical test data; Step 52: Preprocess the temperature time series in the historical experimental data and extract the temperature fluctuation features. Combine these features with the corresponding response time and response effect coefficient to form a multi-dimensional feature vector, which serves as the input feature set for the model. Step 53: Select a deep learning network containing a time series processing module and a fully connected analysis module as the model architecture; divide the input feature set into a training set and a validation set according to the proportion, and use the training set to train the model; Step 54: Use the validation set to evaluate the accuracy and recall performance metrics of the trained model. Once the preset performance requirements are met, solidify the model into an executable security identification model and deploy it to the test bench.

7. The overall distributed control method for the high-gas environment test bench for diesel engines according to claim 6, characterized in that, The historical data includes temperature time series of each authorized component, response time and effect evaluation results of safety components, and manual annotation of the safety level of the corresponding test results under different gas concentrations, temperatures and pressures.

8. The overall distributed control method for the high-gas environment test bench for diesel engines according to claim 6, characterized in that, During the training process, the training set is used to iteratively adjust the model parameters through optimization algorithms to minimize the loss function between the predicted safety level output by the model and the manually labeled safety level.

9. The overall distributed control method for the high-gas environment test bench for diesel engines according to claim 1, characterized in that, The formula for calculating the safety factor in step 6 is as follows: ； In the formula, This represents the safety factor of the final output. Represents the characteristic vector of temperature fluctuation. Represents the temperature fluctuation feature vector of the input. After standardization and normalization, the temperature state safety score is calculated through the sub-network in the safety identification model. Represents a decay function based on response time. Represents the response time of safety components. The coefficient representing the response performance of the safety component. Represents the response time decay factor. Weighting coefficients representing temperature state. The weighting coefficient representing response time. Weighting coefficients representing the response effect.

10. The overall distributed control method for the high-gas environment test bench for diesel engines according to claim 1, characterized in that, The distributed adjustment instructions in step 7 include: When the safety factor is lower than the first threshold, a command is sent to the diesel engine electronic control unit to limit its output power or speed by a preset range. When the safety factor falls below a lower second threshold, a pre-action command is sent directly to the safety component or a higher-level emergency protection procedure is triggered.

Images