A method and system for full-automatic optimization of control performance of a nitrogen-oxygen sensor

Abstract

This invention discloses a fully automatic optimization method and system for the control performance of a nitrogen oxide sensor, comprising the following steps: hardware connection, installing the nitrogen oxide sensor probe onto the gas distribution platform, connecting the sensor controller to a PC via an interface to establish a data acquisition and control channel; writing initial control parameters, writing the initial control parameters to the nitrogen oxide sensor controller via EEPROM or CAN communication; configuring the initial gas distribution environment, controlling the gas distribution platform to introduce a first stable atmosphere via the PC, enabling the nitrogen oxide sensor to operate under known environmental conditions; starting the nitrogen oxide sensor, starting the sensor via CAN communication and waiting for its operating state to stabilize, so that the key controlled variable approaches the target value; and automatically optimizing the control parameters of the nitrogen oxide sensor control system using a model-free optimization algorithm, breaking through the bottleneck of traditional model dependence, adapting to the characteristics of complex nonlinear and strongly coupled control systems, and possessing stronger universality.

Description

Technical Field

[0001] This invention relates to the field of nitrogen and oxygen sensor technology, specifically to a fully automatic method and system for optimizing the control performance of nitrogen and oxygen sensors. Background Technology

[0002] As a core precision instrument in the fields of vehicle exhaust emission monitoring, industrial waste gas treatment, and environmental quality assessment, the control performance of nitrogen oxide sensors directly determines the accuracy and stability of the monitoring system. Existing nitrogen oxide sensors generally employ a multi-level control architecture, such as... Figure 4 and Figure 5 As shown, by adjusting multiple control quantities such as VP0, VP1, and VP2, precise control of controlled quantities such as IP1, V1, and V2 can be achieved.

[0003] However, the control objective and the controlled variable of the nitrogen-oxygen sensor exhibit significant strong coupling and highly nonlinear relationship. Influenced by multiple factors such as the sensor probe material properties, internal structure, and gas diffusion rate, it is difficult to establish an accurate mathematical model. Traditional control optimization methods rely on experienced operators manually adjusting parameters, which has the following significant drawbacks:

[0004] 1. The parameter tuning process is complex and inefficient. Adjusting a single parameter may affect the overall system performance, which is particularly prominent under complex operating conditions.

[0005] 2. The debugging results are highly dependent on human experience, cannot be reproduced, and cannot guarantee that optimal performance will be achieved in the end;

[0006] 3. The control parameters are set offline and cannot be adjusted in real time during system operation. The system performance will degrade significantly after the operating conditions change, and manual readjustment of parameters is required.

[0007] 4. In complex systems with multiple parameters and multiple units, traditional methods require a large number of repeated experiments and data analysis, which is difficult to meet the requirements of sensor response speed and accuracy under dynamic operating conditions.

[0008] Therefore, there is an urgent need for a nitrogen and oxygen sensor control performance optimization method that does not require precise mathematical modeling and can achieve fully automatic online optimization, so as to reduce the dependence on human experience and improve optimization efficiency and system stability. Summary of the Invention

[0009] The purpose of this invention is to provide a fully automatic optimization method and system for the control performance of a nitrogen and oxygen sensor, so as to solve the problems mentioned in the background art.

[0010] To achieve the above objectives, the present invention provides the following technical solution: a fully automatic optimization method for the control performance of a nitrogen and oxygen sensor, comprising the following steps:

[0011] S1. Hardware connection: Install the nitrogen and oxygen sensor probe onto the gas distribution platform, connect the sensor controller to the PC via the interface, and establish a data acquisition and control channel.

[0012] S2. Write the initial control parameters. Write the initial control parameters to the nitrogen and oxygen sensor controller via EEPROM or CAN communication.

[0013] S3. Configure the initial gas mixing environment. Control the gas mixing platform through the PC to introduce the first stable atmosphere so that the nitrogen and oxygen sensor can operate under known environmental conditions.

[0014] S4. Start the nitrogen and oxygen sensor. Start the sensor via CAN communication and wait for its operating status to stabilize so that the key controlled variable is close to the target value.

[0015] S5. Start recording the real-time value of the controlled variable to provide basic data for the evaluation of control performance;

[0016] S6. Apply operating condition disturbance, switch the gas distribution panel to introduce the second stable atmosphere and maintain it for a preset time, then restore to the first stable atmosphere to generate multi-condition response data.

[0017] S7. Control performance evaluation: Calculate the control performance index under the current control parameters. The control performance index is the sum of the absolute values ​​of the deviations between the controlled variable and the corresponding target value;

[0018] S8. Control parameter optimization: Input the current control parameters and performance indicators into the model-free optimization algorithm to calculate a new combination of control parameters. And write it into the sensor controller;

[0019] S9. If the iteration termination condition is met, output the final optimal control parameters; otherwise, return to step S2 to continue iterative optimization.

[0020] Preferably, the nitrogen and oxygen sensor probe achieves precise control of the controlled quantities IP1, V1, and V2 by adjusting the control quantities VP0, VP1, and VP2;

[0021] Steps S7-S8 employ a phased optimization strategy, first optimizing IP1 and V1 control units, then optimizing V2 control unit. The phased optimization strategy specifically involves:

[0022] When optimizing the IP1 and V1 control units, V1 is used as the performance evaluation variable, and only the control parameters corresponding to the IP1 and V1 units are updated.

[0023] When optimizing the V2 control unit, V2 is used as the performance evaluation variable, and only the control parameters corresponding to the V2 unit are updated.

[0024] Preferably, in step S8, the iterative formula of the model-free optimization algorithm is:

[0025] ;

[0026] in, This is the current iteration batch. For the current combination of control parameters, This is the next iteration point, i.e., the new combination of parameters obtained through optimization. This is the current iteration step size. For the iterative gradient.

[0027] Preferred: It gradually decreases as the number of iterations increases, and its calculation formula is:

[0028] ;

[0029] in, The constant coefficients, The deviation coefficient, The convergence index is denoted as .

[0030] Preferred: The calculation method is as follows:

[0031] ;

[0032] in, A set consisting of ±1 randomly generated for each element 3D perturbation vector, The perturbation step size.

[0033] Preferably, in step S9, the iteration termination condition is that the fixed number of iterations reaches a preset upper limit, or that after a preset number of consecutive optimizations, the improvement in the control performance index is lower than a preset threshold.

[0034] Preferably, the first stable atmosphere is compressed air, and the second stable atmosphere is a mixture of 10% O2 and 800ppm NOx.

[0035] An optimized system employing any of the above optimization methods includes:

[0036] The optimization algorithm module is used to generate new control parameters based on the current control parameters and the corresponding control performance indicators, and to determine whether the optimization process meets the termination conditions.

[0037] The controller module is used to receive control parameters output by the optimization algorithm and perform control operations on the nitrogen and oxygen sensor.

[0038] The controlled object module, namely the nitrogen and oxygen sensor, feeds back its state response data to the optimization algorithm module;

[0039] The system disturbance module is used to apply different atmospheric disturbances to the nitrogen and oxygen sensor to verify the robustness of the control system.

[0040] The control performance evaluation module is used to collect and process the response data of the nitrogen and oxygen sensor, calculate the control performance index, and feed it back to the optimization algorithm module.

[0041] Preferably, the optimization algorithm module supports the configuration and switching of multiple model-free optimization algorithms to adapt to different control objectives and performance requirements;

[0042] The controller module supports flexible configuration of various control strategies and structures, enabling precise adjustment of the control quantities VP0, VP1, and VP2 of the nitrogen and oxygen sensors;

[0043] Each module adopts a modular design, which can be independently expanded and combined to realize an automated and repeatable control performance optimization process.

[0044] Compared with the prior art, the beneficial effects of the present invention are:

[0045] 1. The control parameters of the nitrogen and oxygen sensor control system are automatically optimized by a model-free optimization algorithm, which breaks through the bottleneck of traditional model dependence, adapts to the characteristics of complex nonlinear and strongly coupled control systems, and has stronger universality.

[0046] 2. By dynamically optimizing control parameters in real time, the system can adaptively adjust according to different working conditions, resulting in higher response speed and accuracy, and avoiding the performance degradation of traditional offline parameter tuning methods under changing conditions.

[0047] 3. By collecting sensor output data in real time and combining it with a performance evaluation mechanism, control parameters are automatically adjusted. During the optimization and iteration process, online optimization algorithms are continuously used for adjustment until the system performance reaches the predetermined standard. This achieves a fully automated control and optimization process, greatly reducing the need for manual intervention and optimizing the allocation of human resources. Attached Figure Description

[0048] Figure 1 This is a flowchart of the optimization method of the present invention;

[0049] Figure 2 This is a block diagram of the optimized system of the present invention;

[0050] Figure 3 This is a schematic diagram of the hardware connection of the present invention;

[0051] Figure 4 This is a structural diagram of a nitrogen and oxygen sensor probe;

[0052] Figure 5 This is a block diagram of the nitrogen and oxygen sensor control system. Detailed Implementation

[0053] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0054] Please see Figure 4 and Figure 5 As shown, the nitrogen oxide sensor control system achieves precise control of the controlled variables IP1, V1, and V2 by adjusting the control quantities VP0, VP1, and VP2, ensuring that each controlled variable remains near its predetermined target value under different atmospheric conditions. This system has multiple control objectives, and significant coupling and nonlinear relationships exist between different controlled variables and the control quantities. Furthermore, the nitrogen oxide sensor control system is a multi-level control system involving numerous control parameters, increasing the complexity of control performance optimization.

[0055] like Figure 2 As shown, the control performance optimization system of the present invention consists of five major functional modules: optimization algorithm module, controller module, controlled object module, system disturbance module, and control performance evaluation module. Each module works together to achieve fully automatic optimization of control performance.

[0056] 1. Optimization Algorithm Module: This module generates new control parameters based on the current control parameters and their corresponding control performance indicators, and determines whether the optimization process meets the termination conditions. This module supports multiple optimization algorithms to adapt to different control objectives and performance requirements.

[0057] 2. Controller Module: This module receives control parameters output from the optimization algorithm and executes control operations on the controlled object. It can be flexibly configured with different control strategies and structures to achieve precise and adjustable control.

[0058] 3. Controlled Object Module: This is the target system acted upon by the controller, and its state response directly reflects the control effect. State data is fed back to the optimization algorithm module to guide the iterative update of control parameters.

[0059] 4. System Disturbance Module: Apply disturbances to the controlled object to verify the performance and robustness of the control system, and provide multi-condition performance evaluation basis for optimization algorithms.

[0060] 5. Control Performance Evaluation Module: This module collects, processes, and quantifies the response of the controlled object under disturbance conditions, evaluates the control performance, and feeds the results back to the optimization algorithm module to guide subsequent optimization.

[0061] Each module adopts a modular design, which can be flexibly expanded and combined to ensure that the system achieves automated, repeatable and efficient control performance optimization.

[0062] like Figure 1 and Figure 3As shown, the specific flow of the control performance optimization method of the present invention is as follows:

[0063] Step 1: Hardware connection. Install the nitrogen and oxygen sensor probe onto the gas distribution platform, and connect the sensor controller to the PC via the interface to establish a data acquisition and control channel.

[0064] Step 2: Control Parameter Writing: Initial control parameters are written to the nitrogen and oxygen sensor controller via EEPROM or CAN communication to provide an optimization starting point for the control system.

[0065] Step 3: Gas distribution settings. Operate the gas distribution platform on the PC to introduce a first stable atmosphere (such as compressed air) into the gas distribution pipeline to ensure that the controlled object operates under known environmental conditions.

[0066] Step 4: Start the nitrogen and oxygen sensor. Start the nitrogen and oxygen sensor via CAN communication and wait for its operating status to stabilize. At this time, key controlled variables such as IP1, V1, and V2 are close to the target values, which prepares data for subsequent performance evaluation.

[0067] Step 5: Data recording. Begin recording the real-time values ​​of the controlled variable to provide basic data for subsequent control performance evaluation.

[0068] Step 6: Gas distribution setting: Switch the gas distribution line to introduce a second stable atmosphere (e.g., 10% O2 + 800ppm NOx), maintain it for a certain period of time, and then return to the first stable atmosphere to generate multi-condition response data and evaluate the robustness of the control system.

[0069] Step 7: Control performance evaluation. Sum the absolute values ​​of the recorded deviations between the controlled variable and the corresponding target value to obtain the control performance index under the current control parameters. The smaller the indicator, the better the control performance.

[0070] The optimization of the IP1 and V1 control units uses V1 as the performance evaluation variable.

[0071] The optimization of the V2 control unit uses V2 as the performance evaluation variable.

[0072] Step 8: Control parameter optimization. Input the current control parameters and performance indicators into the optimization algorithm to calculate the new combination of control parameters. The algorithm iteration formula is as follows:

[0073] ;

[0074] in, This is the current iteration batch. For the current combination of control parameters, This is the next iteration point, i.e., the new combination of parameters obtained through optimization;

[0075] This is the current iteration step size, which gradually decreases as the number of iterations increases, and can be expressed by the formula... Calculate, where, The constant coefficients, The deviation coefficient, The convergence exponent;

[0076] The gradient is calculated as follows:

[0077] ;

[0078] in, A set consisting of ±1 randomly generated for each element 3D perturbation vector, For perturbation step size, Calculate performance indicators for control purposes;

[0079] Optimize the output by updating only the control parameters of IP1 and V1 units, and only the control parameters of V2 unit.

[0080] Step 9: Iteration Termination. Determine if the iteration termination condition has been met: the fixed number of iterations has reached its upper limit, or the improvement in control performance after multiple consecutive optimizations is lower than a preset threshold. If the condition is met, terminate the optimization and output the final control parameters; otherwise, return to Step 2 and continue iterative optimization.

[0081] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A fully automatic optimization method for the control performance of a nitrogen and oxygen sensor, characterized in that, Includes the following steps: S1. Hardware connection: Install the nitrogen and oxygen sensor probe onto the gas distribution platform, connect the sensor controller to the PC via the interface, and establish a data acquisition and control channel. S2. Write the initial control parameters. Write the initial control parameters to the nitrogen and oxygen sensor controller via EEPROM or CAN communication. S3. Configure the initial gas mixing environment. Control the gas mixing platform through the PC to introduce the first stable atmosphere so that the nitrogen and oxygen sensor can operate under known environmental conditions. S4. Start the nitrogen and oxygen sensor. Start the sensor via CAN communication and wait for its operating status to stabilize so that the key controlled variable is close to the target value. S5. Start recording the real-time value of the controlled variable to provide basic data for the evaluation of control performance; S6. Apply operating condition disturbance, switch the gas distribution panel to introduce the second stable atmosphere and maintain it for a preset time, then restore to the first stable atmosphere to generate multi-condition response data. S7. Control performance evaluation: Calculate the control performance index under the current control parameters. The control performance index is the sum of the absolute values ​​of the deviations between the controlled variable and the corresponding target value; S8. Control parameter optimization: Input the current control parameters and performance indicators into the model-free optimization algorithm to calculate a new combination of control parameters. And write it into the sensor controller; S9. If the iteration termination condition is met, output the final optimal control parameters; otherwise, return to step S2 to continue iterative optimization.

2. The fully automatic optimization method for the control performance of a nitrogen and oxygen sensor according to claim 1, characterized in that: The nitrogen and oxygen sensor probe achieves precise control of the controlled quantities IP1, V1, and V2 by adjusting the control quantities VP0, VP1, and VP2. Steps S7-S8 employ a phased optimization strategy, first optimizing IP1 and V1 control units, then optimizing V2 control unit. The phased optimization strategy specifically involves: When optimizing the IP1 and V1 control units, V1 is used as the performance evaluation variable, and only the control parameters corresponding to the IP1 and V1 units are updated. When optimizing the V2 control unit, V2 is used as the performance evaluation variable, and only the control parameters corresponding to the V2 unit are updated.

3. The fully automatic optimization method for the control performance of a nitrogen and oxygen sensor according to claim 1, characterized in that: In step S8, the iterative formula of the model-free optimization algorithm is: ； in, This is the current iteration batch. For the current combination of control parameters, This is the next iteration point, i.e., the new combination of parameters obtained through optimization. This is the current iteration step size. For the iterative gradient.

4. The fully automatic optimization method for the control performance of a nitrogen and oxygen sensor according to claim 3, characterized in that: It gradually decreases as the number of iterations increases, and its calculation formula is: ； in, The constant coefficients, The deviation coefficient, The convergence index is denoted as .

5. The fully automatic optimization method for the control performance of a nitrogen and oxygen sensor according to claim 3, characterized in that: The calculation method is as follows: ； in, A set consisting of ±1 randomly generated for each element 3D perturbation vector, The perturbation step size.

6. The fully automatic optimization method for the control performance of a nitrogen and oxygen sensor according to claim 1, characterized in that: In step S9, the iteration termination condition is that the fixed number of iterations reaches a preset upper limit, or after a preset number of consecutive optimizations, the improvement in the control performance index is lower than a preset threshold.

7. The fully automatic optimization method for the control performance of a nitrogen and oxygen sensor according to claim 1, characterized in that: The first stable atmosphere is compressed air, and the second stable atmosphere is a mixture of 10% O2 and 800 ppm NOx.

8. An optimization system for implementing the optimization method according to any one of claims 1-7, characterized in that, include: The optimization algorithm module is used to generate new control parameters based on the current control parameters and the corresponding control performance indicators, and to determine whether the optimization process meets the termination conditions. The controller module is used to receive control parameters output by the optimization algorithm and perform control operations on the nitrogen and oxygen sensor. The controlled object module, namely the nitrogen and oxygen sensor, feeds back its state response data to the optimization algorithm module; The system disturbance module is used to apply different atmospheric disturbances to the nitrogen and oxygen sensor to verify the robustness of the control system. The control performance evaluation module is used to collect and process the response data of the nitrogen and oxygen sensor, calculate the control performance index, and feed it back to the optimization algorithm module.

9. The fully automatic optimization system for nitrogen and oxygen sensor control performance according to claim 8, characterized in that: The optimization algorithm module supports the configuration and switching of various model-free optimization algorithms to adapt to different control objectives and performance requirements; The controller module supports flexible configuration of various control strategies and structures, enabling precise adjustment of the control quantities VP0, VP1, and VP2 of the nitrogen and oxygen sensors; Each module adopts a modular design, which can be independently expanded and combined to realize an automated and repeatable control performance optimization process.

Images