A method for optimizing a PID pressure controller

Abstract

The present application relates to pressure controller optimization method technical field, propose a kind of PID pressure controller optimization method, solve the pressure fluctuation in methanol supply system problem.Step is: S100: establish the process dynamic model of marine methanol fuel supply system, clear each key equipment and operating condition;S200: the error of the set value SP and process value PV of methanol fuel supply system PID pressure controller with the integral value of time, construct fitness function;S300: the proportional coefficient and integral coefficient of methanol fuel supply system PID pressure controller are used as optimization variable, constitute the search space of two dimensions in particle swarm algorithm;S400: the dynamic simulation of methanol supply system process is carried out under multiple conditions, and the PID pressure controller performance of marine methanol fuel supply system under different conditions is optimized using particle swarm algorithm;Through multi-objective optimization algorithm adjustment controller parameter;S500: adjustment is carried out based on actual application scene, reduce pressure fluctuation, improve energy efficiency and system stability.

Description

Technical Field

[0001] This invention relates to the technical field of pressure controller optimization methods, specifically to an optimization method for a PID pressure controller in a marine methanol fuel supply system. Background Technology

[0002] In recent years, the marine industry has been committed to replacing traditional high-polluting fuels with clean fuels. Due to methanol's unique advantage of being able to be stored at room temperature among many clean fuels, its application in the marine sector is becoming increasingly widespread. During the operation of methanol-powered ships, when the ship's operating conditions change, the engine load changes, the supply flow rate changes accordingly, and the system pressure fluctuates, causing disturbances to the stable operation of the engine.

[0003] Therefore, to ensure the stable navigation of methanol-powered ships, the performance of the methanol supply system determines the stable and reliable operation of the engine. In existing technologies, system pressure control is typically achieved through a PID controller. The PID controller automatically adjusts the speed of the fuel pumps in the supply system to counteract pressure fluctuations caused by flow rate fluctuations. Optimizing the PID controller parameters is the best way to improve control performance. To optimize the pressure control performance of the methanol supply system and solve the pressure fluctuation problem during operation, optimization algorithms can be used to optimize the control parameters of the PID controller to improve the stability of system pressure.

[0004] The aforementioned PID (Proportional-Integral-Derivative) control is a linear control, consisting of a proportional unit, an integral unit, and a derivative unit. It measures the deviation between the setpoint SP and the process value PV. proportion ,integral ,differential The controlled object is controlled by forming a control quantity through linear combination. The mathematical expression of PID control is as follows:

[0005] ;

[0006] Therefore, the performance of a PID controller depends on the proportional gain. Integral coefficient and differential coefficients Setting three control parameters. Currently, equipment equipped with PID controllers has fixed control parameters set at the factory. This leads to significant fluctuations in the performance of the PID controller in some scenarios, especially in methanol supply systems with multiple operating conditions. Therefore, it is necessary to tune and optimize the PID parameters for different operating conditions to ensure the continuous and stable operation of the methanol supply system. Summary of the Invention

[0007] One of the objectives of this invention is to propose a PID pressure controller optimization method to solve the pressure fluctuation problem in methanol supply systems.

[0008] The technical solution of the present invention is as follows:

[0009] A method for optimizing a PID pressure controller includes the following steps:

[0010] S100: Establish a dynamic process model for the marine methanol fuel supply system and clarify the key equipment and operating conditions.

[0011] S200: The fitness function is constructed by integrating the error of the setpoint SP and process value PV of the PID pressure controller of the methanol fuel supply system over time, in order to evaluate the system stability and response performance.

[0012] S300: The proportional coefficient of the PID pressure controller for the methanol fuel supply system. and integral coefficient As optimization variables, they constitute the two-dimensional search space in the particle swarm optimization algorithm;

[0013] S400: Performs dynamic simulation of the methanol supply system process under multiple operating conditions. Based on the optimization variables and fitness function, it uses the particle swarm optimization algorithm to optimize the performance of the PID pressure controller of the marine methanol fuel supply system under different operating conditions.

[0014] The controller parameters are adjusted by a multi-objective optimization algorithm to ensure the smooth operation of the system under different load and speed changes;

[0015] S500: Obtains optimal PID pressure controller parameters and adjusts them based on actual application scenarios to reduce pressure fluctuations and improve energy efficiency and system stability.

[0016] Furthermore, the fitness function in step S200 is as follows:

[0017] ;

[0018] in, Indicates the current deviation. Indicates the simulation start time. This indicates the simulation end time.

[0019] Furthermore, the particle swarm optimization in step S400 uses Matlab software, and the simulation software Aspen HYSYS is connected to Matlab to achieve real-time data interaction between Aspen HYSYS and Matlab. The steps are as follows:

[0020] S410: Particle swarm optimization algorithm parameter initialization;

[0021] S420: Randomly initialize the particle swarm;

[0022] S430: Calculate the fitness value for each particle;

[0023] S440: Find the optimal values ​​for individuals and groups;

[0024] S450: Update the speed and location of the example;

[0025] S460: Determine whether the termination condition is met;

[0026] S470: If the termination condition is met, output the optimization result;

[0027] If the termination condition is not met, return to step S420.

[0028] Furthermore, step S410 includes:

[0029] S411: Open a specific HYSYS file using Matlab software;

[0030] S412: Set the PID controller parameter values ​​in HYSYS in Matlab software;

[0031] S413: Start HYSYS dynamic simulation;

[0032] S414: Determine if the simulation has ended;

[0033] S415: Stop if the simulation does not end;

[0034] If the simulation ends, Matlab will read the HYSYS error integral value;

[0035] S416: Close the HYSYS file without saving;

[0036] S417: Stop and proceed to step S420.

[0037] Furthermore, the constraints in step S415 are as follows:

[0038] Using the integrator's end time setting in the Aspen HYSYS simulation software as a constraint, if particle parameters cause the system to terminate the simulation prematurely, a penalty function is used to compensate for the unreasonable parameters provided by the particle swarm optimization algorithm. and The population and objective function value that end the simulation early are eliminated.

[0039] Furthermore, the penalty function formula is:

[0040] ;

[0041] in, The simulation ends at time. This is the value of the objective function after penalty.

[0042] Furthermore, the multiple operating conditions in step S400 include several variable load operating conditions, such as acceleration, deceleration, emergency braking, and cruise.

[0043] The beneficial effects of this invention are as follows:

[0044] 1. The system pressure regulation stability and response performance are improved by verifying the results through simulation and optimizing the control parameters of the PID controller using the particle swarm optimization (PSO) algorithm.

[0045] 2. It can avoid the tedious process of traditional manual tuning and improve the adjustment capability and response speed of the control system.

[0046] 3. The established optimization model has strong versatility and scalability, and can be flexibly adapted to different objective functions and variable dimensions, thereby meeting the design requirements of various marine fuel system control strategies. Attached Figure Description

[0047] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0048] Figure 1 This is a flowchart of Example 1;

[0049] Figure 2 This is a schematic diagram of the process flow simulation for Example 1;

[0050] Figure 3 This is a logical framework diagram of the optimization process in Example 1.

[0051] Figure 2 middle:

[0052] 0: Methanol at atmospheric pressure in the fuel tank;

[0053] 1: Error integral value;

[0054] 2: Centrifugal pump outlet logistics

[0055] 3: Heat exchanger tube-side outlet logistics

[0056] 4: Heat exchanger shell-side outlet logistics

[0057] 5: Flow of materials at the heat exchanger shell-side inlet

[0058] 6: Input flow rate without reflux before the centrifugal pump

[0059] 7: Flow rate after the throttle valve of the centrifugal pump

[0060] 8: Flow path after the throttle valve after the heat exchanger

[0061] 9: The material flowing back to the centrifugal pump inlet via the reflux valve after the heat exchanger.

[0062] 10: Logistics before throttling at the main unit

[0063] 11: Centrifugal pump inlet total flow

[0064] 12: Flow after reflux valve

[0065] P-100: A centrifugal pump for pressurizing methanol;

[0066] E-100: Heat exchanger for methanol heating;

[0067] Q-100: Centrifugal pump power

[0068] MIX-100: Rectifier

[0069] TEE-100: Diverter

[0070] VLV-10: Throttling valve from heat exchanger outlet to main unit

[0071] VLV-100: Throttling valve before initial methanol inlet to the centrifugal pump; VLV-101: Throttling valve at the centrifugal pump outlet.

[0072] VLV-102: Heat exchanger outlet throttle valve

[0073] VLV-103: Return valve on the return line;

[0074] OUTLET: Methanol at the main unit inlet;

[0075] PIC-100: Pressure controller, connected to the inverter before the outlet and centrifugal pump;

[0076] TRF-1: Frequency converter;

[0077] TRF-2: Error Integrator;

[0078] SPRDSHT-1: A spreadsheet for transferring data with Matlab. Detailed Implementation

[0079] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0080] Example 1

[0081] like Figures 1 to 3 As shown, this embodiment provides a PID pressure controller optimization method, including the following steps:

[0082] S100: Establish a dynamic process model for the marine methanol fuel supply system and clarify the key equipment and operating conditions.

[0083] In step S100, the dynamic process model of the methanol fuel supply system mainly consists of methanol inlet flow, methanol outlet flow, fuel pump, fuel pump frequency converter, valves, and pressure controller. The process variable source of the pressure controller is the pressure at the methanol outlet. The pressure setpoint SP (Set point) is given according to the corresponding operating conditions. The output target object is the fuel pump frequency converter. The system pressure is controlled by adjusting the speed of the centrifugal pump through the frequency converter. The load change of the host machine can be simulated by defining the flow rate change function at the methanol outlet. Different operating conditions can be tested by changing the function.

[0084] S200: The fitness function is constructed by integrating the error of the set point (SP) and process value (PV) of the PID pressure controller of the methanol fuel supply system over time, in order to evaluate the system stability and response performance.

[0085] The fitness function is as follows:

[0086] ;

[0087] in, Indicates the current deviation. Indicates the simulation start time. This indicates the simulation end time.

[0088] S300: The proportional coefficient of the PID pressure controller for the methanol fuel supply system. and integral coefficient As optimization variables, they constitute the two-dimensional search space in the particle swarm optimization algorithm;

[0089] The objective function is the integral of the difference between the setpoint and the actual value of the PID pressure controller in the methanol supply system over time. The smaller the integral, the better the pressure control effect corresponding to the PID controller parameters set by the particle swarm optimization algorithm.

[0090] Optimize variable proportion coefficient and integral coefficient The upper and lower limits are shown in the table below:

[0091]

[0092] S400: Performs dynamic simulation of the methanol supply system process under multiple operating conditions, including but not limited to several variable load conditions such as acceleration, deceleration, emergency braking, and cruising.

[0093] Then, based on the optimization variables (proportion coefficients) and integral coefficient Using a particle swarm optimization algorithm with a fitness function, the performance of the PID pressure controller of a marine methanol fuel supply system under different operating conditions is optimized.

[0094] Meanwhile, the controller parameters are adjusted through a multi-objective optimization algorithm to ensure the smooth operation of the system under different loads and speed changes.

[0095] Several particles are initialized, each representing a set of PID parameters. The system error integral value returned by Aspen HYSYS is used as the particle fitness. Particles update their velocity and position based on their own historical best position and the swarm's global best position, iterating through the parameters. A maximum number of iterations is set as the termination condition to search for the optimal solution.

[0096] The above particle swarm optimization algorithm was optimized using Matlab software, and the steps are as follows:

[0097] S410: Initialize particle swarm algorithm parameters.

[0098] Specifically, step S410 includes:

[0099] S411: Open a specific HYSYS file using Matlab software;

[0100] S412: Set the PID controller parameter values ​​in HYSYS in Matlab software;

[0101] S413: Start HYSYS dynamic simulation;

[0102] S414: Determine if the simulation has ended;

[0103] Specifically, in the HYSYS dynamic model, it is necessary to set the integrator integration time, i.e. the simulation duration. The simulation duration and termination time are set according to different operating conditions. When the set termination time is reached, the integrator stops, and the simulation ends.

[0104] If the simulation terminates prematurely due to unreasonable given parameters or system failure during the HYSYS model calculation, it is identified as a fault group, and the corresponding data group is discarded through the penalty function in step S415.

[0105] S415: Stop if the simulation does not end;

[0106] If the simulation ends, Matlab will read the HYSYS error integral value;

[0107] The specific constraints in step S415 above are as follows:

[0108] Using the integrator's end time setting in the Aspen HYSYS simulation software as a constraint, if particle parameters cause the system to terminate the simulation prematurely, a penalty function is used to compensate for the unreasonable parameters provided by the particle swarm optimization algorithm. and The population and objective function value that end the simulation early are eliminated.

[0109] Specifically, the penalty function formula is:

[0110] ;

[0111] in, The simulation ends at time. This is the value of the objective function after penalty.

[0112] S416: Close the HYSYS file without saving;

[0113] S417: Stop and proceed to step S420.

[0114] S420: Randomly initialize the particle swarm;

[0115] S430: Calculate the fitness value for each particle;

[0116] S440: Find the optimal values ​​for individuals and groups;

[0117] S450: Update the speed and location of the example;

[0118] S460: Determine whether the termination condition is met;

[0119] Specifically, the particle swarm optimization algorithm will stop the process when the fitness value reaches the minimum error within the specified range, based on the numerical range of the optimization variables set by the operator and the basic settings of the particle swarm optimization algorithm.

[0120] S470: If the termination condition is met, output the optimization result;

[0121] If the termination condition is not met, return to step S420.

[0122] S500: Obtains optimal PID pressure controller parameters and adjusts them based on actual application scenarios to reduce pressure fluctuations and improve energy efficiency and system stability.

[0123] Figure 2The source model established in Aspen HYSYS is mainly used to demonstrate the optimized target methanol supply system process model. In the figure, PIC-100 is the pressure controller, TRF-1 is the centrifugal pump frequency converter, P-100 is the centrifugal pump controlled by the controller, and OUTLET is the methanol flow path to the engine. PIC-100 monitors the pressure of the OUTLET flow path and transmits the control signal to the centrifugal pump through TRF-1 to complete the system pressure control. This invention optimizes the control effect by optimizing the control parameters of PIC-100, reflecting the compatibility between the optimization algorithm and the dynamic model of the methanol supply system.

[0124] Example 2

[0125] The specific usage method of step S400 above is as follows:

[0126] In use, the simulation software Aspen HYSYS is connected to Matlab to achieve real-time data interaction between Aspen HYSYS and Matlab. Matlab can read and assign data to Aspen HYSYS.

[0127] The Matlab handle for calling Aspen HYSYS is as follows:

[0128] Hysys=actxserver('Hysys.Application.V11.0')

[0129] The handle for obtaining folder information is as follows:

[0130] [stat,mess] = fileattrib('F:\1\2\')

[0131] The handle for activating the HYSYS simulation is as follows:

[0132] simcase.invoke('Activate')

[0133] The handle for connecting Matlab to the Aspen HYSYS spreadsheet is as follows:

[0134] spread = hyspread(hy,'SPRDSHT-1')

[0135] The handle for setting the PID controller parameters of Aspen HYSYS via a control spreadsheet is as follows:

[0136] c = hycell(spread,{'A2','A3'})

[0137] hyset(c{1},x(1))

[0138] hyset(c{2},x(2))

[0139] The Matlab handle for calling the particle swarm optimization algorithm is as follows:

[0140] [x,y] = particleswarm(@TEST,2,lb,ub,options);

[0141] The particle swarm optimization fitness value is obtained by retrieving the handle to the value calculated in the Aspen HYSYS spreadsheet, as follows:

[0142] cellB5 = hycell(spread,{'B5'})

[0143] The handle to the particle swarm optimization (PSO) penalty function is as follows:

[0144] t = HySolver.Integrator.CurrentTime.value;

[0145] if HySolver.Integrator.CurrentTime.value<50

[0146] y=(51-t)*1000.

[0147] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A PID pressure controller optimization method, characterized by, Comprising the following steps: S100: Establish a process dynamic model of the marine methanol fuel supply system, and clarify the key equipment and operating conditions; The process dynamic model mainly consists of methanol inlet stream, methanol outlet stream, fuel pump, fuel pump frequency converter, valve and pressure controller, the process variable source of the pressure controller is the pressure of the methanol outlet, the output target object is the fuel pump frequency converter, and the load change of the main engine is simulated by defining the flow change function of the methanol outlet; S200: The integral value of the error of the pressure set value SP and the pressure process value PV of the PID pressure controller of the methanol fuel supply system with respect to time is constructed into a fitness function to evaluate the stability and response performance of the system; S300: set the proportional coefficient of the PID pressure controller of the methanol fuel supply system and the integral coefficient As optimization variables, the search space of two dimensions in the particle swarm algorithm is constituted; S4 00: Dynamic simulation of the methanol supply system process under multiple working conditions, according to the optimization variables and fitness function, the PID pressure controller performance of the marine methanol fuel supply system under different working conditions is optimized by using particle swarm algorithm; The particle swarm algorithm optimization in S400 uses Matlab software, and connects the simulation software Aspen HYSYS and Matlab to realize real-time data interaction between Aspen HYSYS and Matlab, the steps are as follows: S410: Particle swarm algorithm parameter initialization; S410 includes: S411: Open the HYSYS specific file through Matlab software; S412: Set the PID controller parameter value in HYSYS in Matlab software; S413: Start HYSYS dynamic simulation; S414: Determine whether the simulation is finished; S415: If the simulation is not finished, stop; If the simulation is finished, Matlab reads the error integral value of HYSYS; The constraint conditions in S415 are as follows: With the end time of integrator in Aspen HYSYS as the constraint condition, if the particle parameters cause the system to terminate simulation in advance, the unreasonable particle parameters provided by the particle swarm algorithm are punished by using the penalty function and The population that terminates simulation in advance and the objective function value are eliminated. S416: Close the HYSYS file without saving; S417: Stop and enter step S420; S420: Randomly initialize the particle swarm; S430: Calculate the fitness value corresponding to each particle; S440: Find the individual and group optimal values; S450: Update the speed and position of the example; S460: Determine whether the termination condition is met; S470: If the termination condition is met, output the optimization result; If the termination condition is not met, return to step S420; Adjust the controller parameters through the multi-objective optimization algorithm to ensure the smooth operation of the system under different loads and speed changes; S500: Obtain the optimal PID pressure controller parameters, and adjust based on the actual application scenario to reduce pressure fluctuations and improve energy efficiency and system stability.

2. The PID pressure controller optimization method of claim 1, wherein, The fitness function in step S200 is as follows: ； wherein, represents the current error, represents the simulation start time, represents the simulation end time.

3. The PID pressure controller optimization method of claim 2, wherein, The penalty function formula is: ； wherein, is the time to end the simulation, is the penalized objective function value.

4. The PID pressure controller optimization method of claim 1, wherein, The multiple working conditions in step S400 include acceleration, deceleration, emergency braking and cruising.

Images