PIDNN control-based control method for heat dissipation system of aviation piston engine

Abstract

The application relates to a PIDNN control-based aviation piston engine heat dissipation system control method, a double-closed-loop aviation piston engine heat dissipation control system with a liquid-cooled inner ring and an air-cooled outer ring is constructed; the air-cooled outer ring and the liquid-cooled inner ring obtain actual temperature values through temperature sensors, the data obtained by the sensors are processed and calculated by adopting an adaptive flight parameter calculator and a PIDNN controller to generate control signals, and the control signals act on each executing device to adjust the temperature of the heat dissipation system. The PIDNN is fused into a neuron network from a traditional PID control law, has the advantages of the neuron network and the PID control, has the ability of adaptive learning and memory, and overcomes the defect that the adjustment parameters of the traditional PID control method are complicated. The application can effectively control the aviation piston engine heat dissipation system, the heat source temperature cannot be too high or too low, accurate regulation and control of the engine temperature are realized, the reliability and stability of the engine during work are improved, and the safety and stability during flight are ensured.

Description

Technical Field

[0001] This invention relates to the field of thermal management technology for aero-piston engines, specifically a control method for a cooling system of an aero-piston engine based on PIDNN control. Background Technology

[0002] Aviation piston engines offer advantages such as low fuel consumption and low cost when used in medium-altitude long-endurance fixed-wing unmanned aerial vehicles (UAVs). The performance of the engine's cooling system is crucial to its overall operation, reliability, and stability. However, during UAV flight, altitude changes cause nonlinear variations in atmospheric temperature and density. These variable factors can lead to overheating or undercooling of the engine's cooling system. Insufficient cooling in the engine cylinders can cause engine knocking due to increased fuel-air mixture temperature, affecting overall engine safety. Conversely, excessive cooling can result in excessively low temperatures in the combustion chamber and fuel-air mixture, affecting fuel atomization and fuel economy. Therefore, to ensure adequate cooling for aviation piston engines without overcooling them, appropriate thermal management strategies and intelligent control algorithms should be employed to maintain stable engine operating temperatures and achieve precise temperature control.

[0003] To meet the requirements of different mission scenarios for medium-altitude long-endurance fixed-wing UAVs and ensure the stable operation of aircraft piston engines at different altitudes, it is necessary to adopt appropriate heat dissipation methods and strategies to intelligently control the engine heat source temperature, thereby improving the reliability and stability of aircraft piston engines during operation, ensuring the safety and stability of UAV flight while improving fuel economy. Summary of the Invention

[0004] (a) Technical problems to be solved

[0005] To address the shortcomings of existing technologies, this invention provides a control method for the cooling system of an aero-piston engine based on PIDNN control, constructing a dual-closed-loop cooling control system for the aero-piston engine with an inner loop of liquid cooling and an outer loop of air cooling. By adaptively adjusting the PID parameters through a PIDNN controller, the cumbersome parameter adjustments of traditional PID control methods are eliminated. This invention effectively controls the cooling system of the aero-piston engine, preventing the heat source temperature from becoming excessively high or low, achieving precise temperature control of the engine, improving the reliability and stability of engine operation, and ensuring safety and stability during flight.

[0006] (II) Technical Solution

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a control method for a cooling system of an aero-piston engine based on PIDNN control, comprising the following steps:

[0008] 1) Establish a mathematical model relating flight parameters of an aero-piston engine to parameters of a related heat dissipation module;

[0009] 2) Acquire current flight altitude information, ram air temperature, density and wind speed information entering the engine cowling, heat source temperature information of each heat dissipation module of the aircraft piston engine, and relevant aircraft piston engine load status information through relevant sensors arranged in the aircraft piston engine;

[0010] 3) Input the flight parameters measured by the sensors into the adaptive flight parameter calculator and the established mathematical model for calculation to obtain the required heat dissipation of the engine and the required fan speed;

[0011] 4) Set the first preset temperature, the second preset temperature, and the third preset temperature;

[0012] 5) The engine cooling system adjusts its cooling strategy within a preset temperature range based on the current heat source temperature;

[0013] 6) Read the current engine system heat source temperature;

[0014] 7) Set appropriate initial connection weights for the neural network and the learning step size for the connection weights between each layer;

[0015] 8) Input the data calculated by the adaptive parameter calculator and the desired temperature into the PIDNN controller;

[0016] 9) The system error is controlled by the backpropagation algorithm in the PIDNN controller, and K is continuously adjusted by the neural network. p K i K d Three types of PID parameters;

[0017] 10) Input the calculated control signal into the heat dissipation actuator of this system;

[0018] 11) Output the current heat source temperature;

[0019] 12) Update the parameters and return to step 5 until the actual engine temperature reaches the set target temperature.

[0020] Preferably, the control method for the cooling system of an aircraft piston engine based on PIDNN control is characterized in that the engine heat source temperature includes: engine cylinder temperature, cooling water temperature in the liquid-cooled radiator circuit, air temperature at the cabin inlet and outlet, etc., and the engine load includes: engine speed, torque, and flight Mach number, etc.

[0021] Preferably, the control method for the cooling system of an aero-piston engine based on PIDNN control is characterized in that the three preset temperatures are: the first preset temperature is the critical temperature between the maximum opening of the small and large circulation loops of the thermostat, which is 80°C; the second preset temperature is the critical temperature of the maximum large circulation loop of the thermostat, which is 100°C; and the third preset temperature is the maximum required temperature limit of the engine cylinder, which is 120°C.

[0022] Preferably, the control method for the cooling system of an aero-piston engine based on PIDNN control is characterized in that the engine cooling system strategy is:

[0023] 1) When the coolant temperature is less than 80℃, the liquid-cooled radiator in the heat dissipation system is in a small circulation state under the action of the thermostat.

[0024] 2) When the coolant temperature is greater than or equal to 80℃ and less than 100℃, the thermostat valve adjusts its opening according to the water temperature, and the liquid-cooled radiator starts the large circulation state.

[0025] 3) When the coolant temperature is greater than or equal to 100℃, the thermostat valve reaches its maximum opening. At this time, the cooling fan turns on and adjusts its speed according to the heat source temperature, ensuring that the coolant temperature and cylinder temperature do not exceed 120℃.

[0026] Preferably, the control method for the cooling system of an aero-piston engine based on PIDNN control is characterized in that the neural network in the PIDNN controller adopts a 2-3-1 structure, which is divided into an input layer, a hidden layer, and an output layer; the input layer of the PIDNN controller has two neurons, namely the original set signal value input and the value fed back by the input signal, i.e., the fan speed setpoint N. i and fan speed feedback value N r The hidden layer of the PIDNN controller has three neurons: a proportional neuron, an integral neuron, and a differential neuron. The output layer of the PIDNN controller has one neuron, whose input value is the weighted average of the output values ​​of all hidden layer neurons. The output value is the fan speed N. o ;

[0027] Forward PIDNN algorithm:

[0028] 1) Input layer algorithm

[0029] The two input neurons of the input layer are set to a value N. i With output feedback value N r At any sampling time k, the input function of the neuron in the input layer is:

[0030]

[0031] The state function of the input layer neuron is:

[0032]

[0033] The output function of the input layer neuron is:

[0034]

[0035] In the formula, i represents the sequence number of the input layer neuron, i = 1, 2;

[0036] 2) Hidden Layer Algorithm

[0037] The hidden layer consists of three neurons: a proportional neuron, an integral neuron, and a differential neuron. These three neurons represent P, I, and D functions, respectively. At any sampling time k, its input function is:

[0038]

[0039] In the formula, i represents the neuron number in the input layer, i = 1, 2; j represents the neuron number in the hidden layer, j = 1, 2, 3; w ii This represents the link weights between the input layer and the hidden layer;

[0040] The state function of the proportional element P is:

[0041] u'1(k)=net'1(k)

[0042] The state function of integrator I is:

[0043] u′2(k)=u′2(k-1)+net′2(k)

[0044] The state function of the differential element D is:

[0045] u′3(k)=net′3(k)-net′3(k-1)

[0046] The output function of the hidden layer neurons is:

[0047]

[0048] 3) Output layer algorithm

[0049] The output layer has only one neuron, which is the output of the entire PIDNN network; the input function of the output layer is:

[0050]

[0051] Its state function is:

[0052] u″(k)=net″(k)

[0053] Its output function is:

[0054]

[0055] The PIDNN output value is:

[0056] N o (k)=x″(k)

[0057] PIDNN backpropagation algorithm:

[0058] Objective function E:

[0059]

[0060] In the formula, N i (k) represents the setpoint speed of the cooling fan at time k; N r (k) represents the cooling fan output feedback value at time k; l represents the number of sampling points; e represents the error between the set speed and the actual speed;

[0061] The formula for modifying the link weights from the hidden layer to the output layer is:

[0062]

[0063] In the formula, i represents the sequence number of the hidden layer neuron, i = 1, 2, 3; w i Indicates the link weights from the hidden layer to the output layer; η i The learning step size; l is the number of sampling points; x′ i (k) represents the output value of the i-th input layer neuron;

[0064] The formula for modifying the link weights from the input layer to the hidden layer is:

[0065]

[0066] In the formula, i represents the sequence number of the input layer neuron, i = 1, 2; j represents the sequence number of the hidden layer neuron, j = 1, 2, 3; w ij Indicates the link weights from the input layer to the hidden layer; η i The learning step size; l is the number of sampling points; x′ i (k) represents the output value of the i-th input layer neuron;

[0067] Where the gradient is, δ j (k) is the Dirac function; sgn() is the step function; u′ j (k) is the hidden layer state of the j-th state at time k; net′ j (k) is the input of the j-th hidden layer at time k.

[0068] Preferably, the control method for the cooling system of an aircraft piston engine based on PIDNN control is characterized in that the adaptive flight parameter calculator calculates the parameters to be substituted into the PIDNN control model based on the data measured by the sensors input into the mathematical model; the mathematical model in the adaptive flight parameter calculator is:

[0069] 1) The relationship between the cooling air demand and the heat dissipation of the cooling system can be obtained through the heat balance equation:

[0070]

[0071] In the formula, q a It is the cooling airflow rate; Q w This is the heat dissipation required by the cooling system; C p It is the specific heat capacity of cooling air at constant pressure; ρ a It is the density of the cooling air; Δt a It is the temperature difference between the inlet and outlet air of the heat dissipation system;

[0072] 2) Relationship between fan speed and cooling airflow:

[0073]

[0074] In the formula, N i β0, β1, and β2 are the constant, linear term, and quadratic term coefficients of the linear regression, respectively.

[0075] 3) Calculation of air temperature difference between the inlet and outlet of the heat dissipation system:

[0076] Δt a =t o -t e

[0077] In the formula, t e It is the air temperature at the inlet of the cooling system; t o It is the air temperature at the outlet of the heat dissipation system;

[0078] 4) Relationship between heat absorbed by air and mass flow rate:

[0079] Q a =q m C p Δt a

[0080] In the formula, Q a It is the heat absorbed by the air; q m It is the mass flow rate of air; C p It is the specific heat capacity of air at constant pressure; Δt a It's the rise in air temperature;

[0081] 5) Airflow mass calculation:

[0082] q m =q a ρ a

[0083] In the formula, q a It is the volumetric flow rate of air; ρ a It is the density of ambient air;

[0084] 6) Calculation of stamping air temperature:

[0085] t r =t a +0.2Ma 2 T a

[0086] In the formula, t r t represents the temperature of the press air. a It represents the current ambient air temperature; Ma is the Mach number; T a It is the absolute temperature of the ambient air;

[0087] 7) Calculation of ambient air density:

[0088]

[0089] In the formula, ρ a This is the current ambient air density; t a This is the current ambient air temperature; P a It is the current atmospheric pressure.

[0090] (III) Beneficial Effects

[0091] Compared with the prior art, the present invention provides a control method for the cooling system of an aero-piston engine based on PIDNN control, which has the following advantages:

[0092] 1. The control method provided by the present invention controls the temperature of the heat dissipation system through a PIDNN controller. The neural network structure in the PIDNN has the ability to learn and remember adaptively, and can adaptively adjust the parameters in the PID, thus overcoming the shortcomings of the traditional PID control method, which is cumbersome in adjusting parameters.

[0093] 2. The control method provided by this invention, through the combined effect of thermal management strategies and control algorithms, can effectively control the temperature of the cooling system of an aircraft piston engine, ensuring that the heat source temperature is neither too high nor too low, and achieving precise regulation of the engine temperature. Attached Figure Description

[0094] Figure 1 This invention relates to a dual closed-loop cooling control system for aero-piston engines.

[0095] Figure 2 This is a structural diagram of the PIDNN algorithm of this invention.

[0096] Figure 3 This is a diagram illustrating the heat dissipation strategy of the present invention.

[0097] Figure 4 This is a curve diagram showing the adaptive adjustment of PID parameters in this invention.

[0098] Figure 5 This is a graph showing the fan speed controlled by the PIDNN algorithm of this invention. Detailed Implementation

[0099] 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 embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0100] 1. In summary, combining Figure 1 The control method for the cooling system of an aero-piston engine based on PIDNN control includes the following steps:

[0101] 1) Establish a mathematical model relating flight parameters of an aero-piston engine to parameters of a related heat dissipation module;

[0102] 2) Acquire current flight altitude information, ram air temperature, density and wind speed information entering the engine cowling, heat source temperature information of each heat dissipation module of the aircraft piston engine, and relevant aircraft piston engine load status information through relevant sensors arranged in the aircraft piston engine;

[0103] 3) Input the flight parameters measured by the sensors into the adaptive flight parameter calculator and the established mathematical model for calculation to obtain the required heat dissipation of the engine and the required fan speed;

[0104] 4) Set the first preset temperature, the second preset temperature, and the third preset temperature;

[0105] 5) The engine cooling system adjusts its cooling strategy within a preset temperature range based on the current heat source temperature;

[0106] 6) Read the current engine system heat source temperature;

[0107] 7) Set appropriate initial connection weights for the neural network and the learning step size for the connection weights between each layer;

[0108] 8) Input the data calculated by the adaptive parameter calculator and the desired temperature into the PIDNN controller;

[0109] 9) The system error is controlled by the backpropagation algorithm in the PIDNN controller, and K is continuously adjusted by the neural network. p K i K d Three types of PID parameters;

[0110] 10) Input the calculated control signal into the heat dissipation actuator of this system;

[0111] 11) Output the current heat source temperature;

[0112] 12) Update the parameters and return to step 5 until the actual engine temperature reaches the set target temperature.

[0113] 2. In summary, the control method for the cooling system of an aircraft piston engine based on PIDNN control includes engine heat source temperatures such as engine cylinder temperature, cooling water temperature in the liquid-cooled radiator circuit, and air temperature at the cabin inlet and outlet, and engine loads such as engine speed, torque, and flight Mach number.

[0114] 3. In summary, the control method for the cooling system of an aero-piston engine based on PIDNN control has three preset temperatures: the first preset temperature is the critical temperature between the maximum opening of the small and large circulation loops of the thermostat, which is 80°C; the second preset temperature is the critical temperature of the maximum large circulation loop of the thermostat, which is 100°C; and the third preset temperature is the maximum required temperature limit of the engine cylinder, which is 120°C.

[0115] 4. In summary, combining Figure 3 The control method for the cooling system of an aero-piston engine based on PIDNN control has the following engine cooling system strategy:

[0116] 1) When the coolant temperature is less than 80℃, the liquid-cooled radiator in the heat dissipation system is in a small circulation state under the action of the thermostat.

[0117] 2) When the coolant temperature is greater than or equal to 80℃ and less than 100℃, the thermostat valve adjusts its opening according to the water temperature, and the liquid-cooled radiator starts the large circulation state.

[0118] 3) When the coolant temperature is greater than or equal to 100℃, the thermostat valve reaches its maximum opening. At this time, the cooling fan turns on and adjusts its speed according to the heat source temperature, ensuring that the coolant temperature and cylinder temperature do not exceed 120℃.

[0119] 5. In summary, combining Figure 2 This paper presents a control method for the cooling system of an aero-piston engine based on PIDNN control. The neural network in the PIDNN controller adopts a 2-3-1 structure, which is divided into an input layer, a hidden layer, and an output layer. The input layer of the PIDNN controller has two neurons: the original set signal value and the feedback value of the input signal, i.e., the fan speed setpoint N. i and fan speed feedback value N r The hidden layer of the PIDNN controller has three neurons: a proportional neuron, an integral neuron, and a differential neuron. The output layer of the PIDNN controller has one neuron, whose input value is the weighted average of the output values ​​of all hidden layer neurons. The output value is the fan speed N. o ;

[0120] Forward PIDNN algorithm:

[0121] 1) Input layer algorithm

[0122] The two input neurons of the input layer are set to a value N. i With output feedback value N r At any sampling time k, the input function of the neuron in the input layer is:

[0123]

[0124] The state function of the input layer neuron is:

[0125]

[0126] The output function of the input layer neuron is:

[0127]

[0128] In the formula, i represents the sequence number of the input layer neuron, i = 1, 2;

[0129] 2) Hidden Layer Algorithm

[0130] The hidden layer consists of three neurons: a proportional neuron, an integral neuron, and a differential neuron. These three neurons represent P, I, and D functions, respectively. At any sampling time k, its input function is:

[0131]

[0132] In the formula, i represents the neuron number in the input layer, i = 1, 2; j represents the neuron number in the hidden layer, j = 1, 2, 3; w ij This represents the link weights between the input layer and the hidden layer;

[0133] The state function of the proportional element P is:

[0134] u′1(k)=net′1(k)

[0135] The state function of integrator I is:

[0136] u′2(k)=u′2(k-1)+net′2(k)

[0137] The state function of the differential element D is:

[0138] u′3(k)=net′3(k)-net′3(k-1)

[0139] The output function of the hidden layer neurons is:

[0140]

[0141] 3) Output layer algorithm

[0142] The output layer has only one neuron, which is the output of the entire PIDNN network; the input function of the output layer is:

[0143]

[0144] Its state function is:

[0145] u″(k)=net″(k)

[0146] Its output function is:

[0147]

[0148] The PIDNN output value is:

[0149] N o (k)=x″(k)

[0150] PIDNN backpropagation algorithm:

[0151] Objective function E:

[0152]

[0153] In the formula, N i (k) represents the setpoint speed of the cooling fan at time k; N r (k) represents the cooling fan output feedback value at time k; l represents the number of sampling points; e represents the error between the set speed and the actual speed;

[0154] The formula for modifying the link weights from the hidden layer to the output layer is:

[0155]

[0156] In the formula, i represents the sequence number of the hidden layer neuron, i = 1, 2, 3; w i Indicates the link weights from the hidden layer to the output layer; η i The learning step size; l is the number of sampling points; x′ i (k) represents the output value of the i-th input layer neuron;

[0157] The formula for modifying the link weights from the input layer to the hidden layer is:

[0158]

[0159] In the formula, i represents the sequence number of the input layer neuron, i = 1, 2; j represents the sequence number of the hidden layer neuron, j = 1, 2, 3; w ij Indicates the link weights from the input layer to the hidden layer; η i The learning step size; l is the number of sampling points; x′ i (k) represents the output value of the i-th input layer neuron;

[0160] Where the gradient is, δ j (k) is the Dirac function; sgn() is the step function; u′ j (k) is the hidden layer state of the j-th state at time k; net′ j (k) is the input of the j-th hidden layer at time k.

[0161] 6. In summary, this PIDNN-based control method for the cooling system of an aircraft piston engine uses an adaptive flight parameter calculator to calculate the parameters to be substituted into the PIDNN control model based on data measured by sensors. The environment affects forced convection heat transfer by influencing the physical properties of air. The main environmental factors include air temperature, pressure, and humidity. Thermal conductivity, specific heat capacity, density, and viscosity also change with these environmental factors.

[0162] The mathematical model in the adaptive flight parameter calculator is:

[0163]

[0164] 1) The relationship between the cooling air demand and the heat dissipation of the cooling system can be obtained through the heat balance equation:

[0165]

[0166] In the formula, q a It is the cooling airflow rate; Q w This is the heat dissipation required by the cooling system; C p It is the specific heat capacity of cooling air at constant pressure; ρ a It is the density of the cooling air; Δt a It is the temperature difference between the inlet and outlet air of the heat dissipation system;

[0167] 2) Relationship between fan speed and cooling airflow:

[0168]

[0169] In the formula, N i β0, β1, and β2 are the constant, linear term, and quadratic term coefficients of the linear regression, respectively.

[0170] The linear regression equation needs to be established based on the resistance-flow matching curve of the cooling fan and the radiator. For specific methods, refer to GB / T 1236-2017 Standardized Air Duct Performance Test for Industrial Ventilation Fans.

[0171] 3) Calculation of air temperature difference between the inlet and outlet of the heat dissipation system:

[0172] Δt a =t o -t e

[0173] In the formula, t e It is the air temperature at the inlet of the cooling system; t o It is the air temperature at the outlet of the heat dissipation system;

[0174] The inlet and outlet temperatures of the heat dissipation system are measured by the installed temperature sensors;

[0175] 4) Relationship between heat absorbed by air and mass flow rate:

[0176] Q a =q m C p Δt a

[0177] In the formula, Q a It is the heat absorbed by the air; q m It is the mass flow rate of air; C p It is the specific heat capacity of air at constant pressure; Δt a It's the rise in air temperature;

[0178] 5) Airflow mass calculation:

[0179] q m =q a ρ a

[0180] In the formula, q a It is the volumetric flow rate of air; ρ a It is the density of ambient air;

[0181] 6) Calculation of stamping air temperature:

[0182] t r =t a +0.2Ma 2 T a

[0183] In the formula, t r t represents the temperature of the press air. a It represents the current ambient air temperature; Ma is the Mach number; T a It is the absolute temperature of the ambient air;

[0184] 7) Calculation of ambient air density:

[0185]

[0186] In the formula, ρ a This is the current ambient air density; t a This is the current ambient air temperature; P a This refers to the current atmospheric pressure.

[0187] Substituting the mathematical model established above into the adaptive flight parameter calculator, the required system parameters, such as the required heat dissipation of the system and the corresponding required speed of the cooling fan, are obtained under different altitudes and flight conditions.

[0188] 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 control method for a cooling system of an aero-piston engine based on PIDNN control, characterized in that, Includes the following steps: (1) Establish a mathematical model relating the flight parameters of an aero-piston engine to the parameters of a related heat dissipation module; The mathematical model is as follows: 1) The relationship between the cooling air demand and the heat dissipation of the cooling system can be obtained through the heat balance equation: In the formula, It is the cooling airflow; This is the amount of heat dissipated by the cooling system; It is the specific heat capacity of cooling air at constant pressure; It is the ambient air density; It is the temperature difference between the inlet and outlet air of the heat dissipation system; 2) Relationship between fan speed and cooling airflow: In the formula, It refers to the fan speed; These are the constant, linear term, and quadratic term coefficients of the linear regression, respectively. 3) Calculation of air temperature difference between the inlet and outlet of the heat dissipation system: In the formula, It is the air temperature at the inlet of the heat dissipation system; It is the air temperature at the outlet of the heat dissipation system; 4) Relationship between heat absorbed by air and mass flow rate: In the formula, It is the heat absorbed by the air; It is the mass flow rate of air; It is the specific heat capacity of air at constant pressure; It is the temperature difference of the air; 5) Airflow mass calculation: In the formula, It is the airflow rate; It is the density of ambient air; 6) Calculation of stamping air temperature: In the formula, The temperature of the press air; This is the current ambient air temperature; It is the Mach number; It is the absolute temperature of the ambient air; 7) Calculation of ambient air density: In the formula, This is the current ambient air density; This is the current ambient air temperature; This refers to the current atmospheric pressure. (2) Obtain information on current flight altitude, temperature, density and wind speed of ram air entering the engine cowling, heat source temperature of each heat dissipation module of the aircraft piston engine, and relevant aircraft piston engine load and operating status information through relevant sensors arranged in the aircraft piston engine; (3) Input the flight parameters measured by the sensors into the adaptive flight parameter calculator and the established mathematical model for calculation to obtain the required heat dissipation of the engine and the required speed of the fan; (4) Set the first preset temperature, the second preset temperature and the third preset temperature; the three preset temperatures are as follows: the first preset temperature is the critical temperature between the maximum small circulation and the large circulation of the thermostat, which is 80°C; the second preset temperature is the critical temperature of the maximum large circulation of the thermostat, which is 100°C; and the third preset temperature is the maximum limit temperature required by the engine cylinder, which is 120°C. (5) The engine cooling system adjusts its cooling strategy within a preset temperature range based on the current heat source temperature; The engine cooling system strategy is as follows: 1) When the coolant temperature is less than 80℃, the liquid-cooled radiator in the heat dissipation system is in a small circulation state under the action of the thermostat. 2) When the coolant temperature is greater than or equal to 80℃ and less than 100℃, the thermostat valve adjusts its opening according to the water temperature, and the liquid-cooled radiator starts the large circulation state. 3) When the coolant temperature is greater than or equal to 100℃, the thermostat valve reaches its maximum opening. At this time, the cooling fan turns on and adjusts its speed according to the heat source temperature, ensuring that the coolant temperature and cylinder temperature do not exceed 120℃. (6) Read the current engine system heat source temperature; (7) Set appropriate initial connection weights for the neural network and learning step sizes for the connection weights between layers; (8) Input the data calculated by the adaptive flight parameter calculator and the desired temperature into the PIDNN controller; (9) The system error is controlled by the backpropagation algorithm in the PIDNN controller and continuously adjusted by the neural network. , , Three types of PID parameters; (10) Input the calculated control signal into the heat dissipation actuator of this system; (11) Output the current heat source temperature; (12) Update the parameters and return to step (5) until the actual engine temperature reaches the set target temperature.

2. The control method for a cooling system of an aero-piston engine based on PIDNN control according to claim 1, characterized in that, The engine heat source temperatures include: engine cylinder temperature, cooling water temperature in the liquid-cooled radiator circuit, and air temperature at the engine compartment inlet and outlet; engine load includes torque; engine operating parameters include engine speed and flight Mach number.

Images