Model predictive temperature control method and system for air-cooled heating process

Abstract

The application discloses a model prediction temperature control method and system for air cooling heating process, and relates to the field of model prediction temperature control. The method identifies a heating model of the heating process, predicts inertia change of future temperature in an integral type increment model mode, so that the control quantity can be reduced in advance before the heating process reaches a set temperature. The control quantity and control quantity change constraint are introduced simultaneously in the heating process. The control quantity weight matrix and control quantity change weight coefficient are introduced in the controller design. The application can ensure that the control quantity is always greater than or equal to 0 in the temperature regulation process, and the temperature regulation can reach the set value in a gentle mode to realize smooth temperature regulation, no overshoot and no oscillation. The application can be widely applied to the field of temperature control of air cooling heating process, and has important practical value.

Description

Technical Field

[0001] This invention relates to the field of air-cooled heating temperature control, and in particular to a model-predictive temperature control method and system for air-cooled heating processes. Background Technology

[0002] Temperature control is a typical industrial control problem, widely used in thermal power generation, chemical processes, and CNC systems. Temperature regulation must adhere to the principles of energy conservation and heat exchange, exhibiting characteristics such as large time delays and large inertia. Due to strict physical constraints, it is often difficult to simultaneously achieve optimal performance in temperature control, such as speed, overshoot, and stability, directly impacting the quality and efficiency of industrial processes. Therefore, temperature control has become a challenging problem in control theory research.

[0003] Air-cooled heating is a crucial aspect of temperature control, and its control methods and technologies are constantly evolving and improving. Currently, PID controllers are the most widely used type of temperature controller in industrial systems. PID controllers possess strong robustness and can achieve wide-range temperature control and regulation without requiring a precise model of the temperature process. Early research focused on typical first- and second-order temperature processes, presenting frequency-domain and time-domain design methods for PID controllers, suitable for controlling temperature processes with time-delay characteristics. To achieve high-precision temperature control, PID controllers have been further combined with various advanced control strategies, forming typical temperature control schemes such as multi-degree-of-freedom PID control, Smith predictor control, and internal model PID control. These schemes, primarily PID-driven, have a profound impact on process control.

[0004] In recent years, with the development of advanced control theory, temperature control technology has also been gradually improved. Some new temperature control systems employ advanced sensors and control technologies to achieve more precise temperature control. Compared with PID control, although the theoretical design of advanced control strategies is more complex, it can improve control performance and overcome the limitations of linear control through time-varying and nonlinear feedback adjustment. However, conventional temperature control strategies still face the problem of failing to achieve good control results in air-cooled systems. Summary of the Invention

[0005] The purpose of this invention is to provide a model-predictive temperature control method and system for air-cooled heating processes, which can achieve temperature regulation to reach the set value in a smooth manner, overcome large abrupt changes in control variables, and achieve stable temperature regulation without overshoot or oscillation, thus solving the problem that conventional temperature control strategies are difficult to achieve good control effects in air-cooled systems.

[0006] To achieve the above objectives, the present invention provides the following solution:

[0007] A model-predictive temperature control method for an air-cooled heating process, the method comprising:

[0008] Identify the heating model of the controlled object in the air-cooled heating process, and determine the steady-state gain and time constant in the heating model;

[0009] The heating model is transformed into an integral incremental model. The heating state of the heating model is predicted using the integral incremental model to obtain the predicted heating state. The predicted temperature output of the integral incremental model is determined using the integral incremental model and the predicted heating state.

[0010] Determine the control quantity constraints and control quantity variation constraints of the heating model; the upper limit of the control quantity variation constraint is determined based on the steady-state gain, the time constant, and the upper bound coefficient of the control quantity variation.

[0011] A control quantity change weight coefficient is introduced, and the values ​​of the control quantity change weight coefficient and the control quantity change upper bound coefficient are determined based on the overshoot and oscillation degree of the current heating model.

[0012] The objective function for solving the control quantity change is determined based on the reference temperature input of the heating model, the predicted temperature output of the integral incremental model, and the control quantity change weighting coefficient.

[0013] The current control quantity change is solved based on the objective function of the control quantity change and the control quantity constraints and control quantity change constraints. The current control quantity of the heating model is determined based on the current control quantity change, and the heating model is temperature controlled according to the current control quantity.

[0014] Optionally, the expression for the heating model is:

[0015]

[0016] Where K is the steady-state gain; T is the time constant; ζ is the damping coefficient; y(s) is the temperature output; u(s) is the control input; and s is the frequency domain operator.

[0017] Optionally, the expression for the integral incremental model is:

[0018]

[0019] Among them, A m B m C m The system matrix after transforming the heating model G(s) into a discrete state-space model; Δx m(k) represents the change of state variables after transforming the heating model G(s) into a discrete state-space model at time k; A, B, and C are the system matrices after transforming the discrete state-space model into an integral-increment model; x(k) represents the state variables after transforming the discrete state-space model into an integral-increment model; Δu(k) represents the change of control quantity at time k; o m y(k) is a zero matrix; y(k) represents the output temperature of the integral incremental model at time k.

[0020] Optionally, the prediction expression for the predicted heating state is:

[0021]

[0022] Where x(k+N) p ) is the Nth prediction of the current heating state variable x(k) p One heating state; N p For the given prediction time domain; N c For a given control time domain, N c <N p The change of the control quantity outside the control time domain is set to zero; Δu(k), Δu(k+1), ... Δu(k+N) c -1) represents the future N that needs to be calculated. c Changes in the control quantity.

[0023] Optionally, the expression for the predicted temperature output of the integral incremental model is:

[0024] Y=Fx(k)+φΔU

[0025] Among them, Y=[y(k+1),y(k+2),…y(k+N p )] T ;

[0026] ΔU=[Δu(k),Δu(k+1),…Δu(k+N c -1)] T ;

[0027]

[0028] In the formula, y(k+1), y(k+2), ..., y(k+N) p ) represents the predicted N p Temperature output.

[0029] Optionally, the formula for calculating the upper limit of the control quantity change in the control quantity change constraint is as follows:

[0030] Δu max =k*T / K

[0031] Where, Δumax k is the upper limit of the control quantity's change. * >0 represents the upper bound coefficient for the change in the control quantity.

[0032] Optionally, the values ​​of the control quantity change weight coefficient and the control quantity change upper bound coefficient are determined based on the overshoot and oscillation degree of the current heating model, specifically including:

[0033] The control quantity change weight coefficient and the control quantity change upper bound coefficient are adjusted based on the identity matrix or a preset value.

[0034] When the overshoot is greater than a first preset value or the oscillation is greater than a second preset value, the current control quantity change weight coefficient is increased and the current control quantity change upper bound coefficient is decreased; conversely, the current control quantity change weight coefficient is decreased and the current control quantity change upper bound coefficient is increased.

[0035] Optionally, the formula for calculating the reference temperature input of the heating model is:

[0036]

[0037] Among them, y s y0 is the reference temperature input; y0 is the initial temperature; y * ω is the desired temperature. c It represents excessive speed.

[0038] Optionally, the expression for the objective function of solving the control variable change is:

[0039]

[0040] in, Let be the control variable weighting coefficient matrix; the first element of the solved ΔU is used for the current control variable update, Δu(k) = [1 0 … 0]ΔU; y s The generated temperature sequence is denoted as

[0041] This invention provides a model predictive temperature control system for an air-cooled heating process, the system comprising:

[0042] The heating model identification module is used to identify the heating model of the controlled object in the air-cooled heating process and determine the steady-state gain and time constant in the heating model.

[0043] The model prediction module is used to convert the heating model into an integral incremental model, use the integral incremental model to predict the heating state of the heating model, obtain the predicted heating state, and use the integral incremental model and the predicted heating state to determine the predicted temperature output of the integral incremental model.

[0044] The constraint construction module is used to determine the control quantity constraints and control quantity variation constraints of the heating model; the upper limit value of the control quantity variation constraint is determined based on the steady-state gain, the time constant, and the upper bound coefficient of the control quantity variation.

[0045] The weight coefficient determination module is used to introduce the control quantity change weight coefficient and determine the value of the control quantity change weight coefficient and the value of the control quantity change upper bound coefficient based on the overshoot and oscillation degree of the current heating model.

[0046] The objective function construction module is used to determine the objective function for solving the control quantity change based on the reference temperature input of the heating model, the predicted temperature output of the integral incremental model, and the control quantity change weight coefficient.

[0047] The temperature control module is used to solve the current control quantity change based on the objective function of the control quantity change and the control quantity constraints and control quantity change constraints, determine the current control quantity of the heating model based on the current control quantity change, and perform temperature control on the heating model according to the current control quantity.

[0048] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0049] This invention provides a model-predictive temperature control method and system for air-cooled heating processes. By identifying a second-order model of the heating process and predicting future temperature inertial changes using an incremental model, the control variable can be reduced in advance before the set temperature is reached, thus achieving more precise control of the heating process. Constraints on the control variable and its changes are introduced to avoid drastic temperature changes caused by sudden changes in the control variable during heating, ensuring the stability and accuracy of the control process. Incremental models, control variable constraints, and control variable change constraints are incorporated into the controller design to achieve precise control of the heating process. Control variable weighting coefficients and control variable change upper bound coefficients are introduced; by adjusting these two types of coefficients, the temperature of the heating process can be controlled more precisely. This invention enables temperature regulation to reach the set value smoothly, overcoming large abrupt changes in the control variable, achieving stable temperature regulation without overshoot or oscillation, and solving the problem that conventional temperature control strategies are difficult to achieve good control effects in air-cooled systems. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 This is a schematic diagram of the temperature prediction and control structure for the air-cooled heating process provided in Embodiment 1 of the present invention.

[0052] Figure 2 A flowchart of a model-predicted temperature control method for an air-cooled heating process provided in Embodiment 1 of the present invention;

[0053] Figure 3 The model prediction smooth temperature control method for the air-cooled heating process in the second-order system provided in Embodiment 1 of the present invention has different control increment constraint control effects.

[0054] Figure 4 The model prediction smooth temperature control method for the air-cooled heating process in the second-order system provided in Embodiment 1 of the present invention has different control weights to show the control effect.

[0055] Figure 5 The comparison between model predictive control and PID control provided in Embodiment 1 of the present invention. Detailed Implementation

[0056] 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.

[0057] The purpose of this invention is to provide a model-predictive temperature control method and system for air-cooled heating processes, which can achieve temperature regulation to reach the set value in a smooth manner, overcome large abrupt changes in control variables, and achieve stable temperature regulation without overshoot or oscillation, thereby solving the temperature control problem in air-cooled heating processes.

[0058] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0059] Example 1

[0060] This embodiment provides a model-predictive temperature control method for an air-cooled heating process, such as... Figure 1As shown, this diagram illustrates the principle structure of a temperature predictive control with constrained control quantity and control increment during the heating phase. This control principle structure includes components such as air-cooling gradual temperature generation, integral incremental model temperature prediction, real-time constraint optimization of the control quantity, and adjustment of the control quantity and its variation coefficients. The four components enable gradual temperature regulation at different levels. Figure 1 This system combines four components to achieve smooth temperature regulation during the air-cooled heating process. The first component generates a suitable temperature setpoint, ensuring it doesn't jump. The second component predicts the inertial temperature changes (heating or air-cooling) over a future period, preventing oscillations and overshoot in the system due to the current control input. The third component optimizes real-time control constraints, controlling jumps in the control input and its increments. The fourth component adjusts the coefficients of the control input and its changes, introducing a quantitative adjustment mechanism to constrain new control inputs. This design achieves smooth temperature control, avoiding overshoot and oscillations, and reducing environmental impact.

[0061] The temperature control system of the 3D printer extruder head is a typical air-cooled heating process, and the effectiveness of extruder head temperature control determines the quality of the printed product. The printer uses a single-head heating element, model 12V, 40W, and temperature measurement uses a thermistor, model NTC-R25=100K±1%B25 / 50=3950. The thermistor contains three important parameters: reference temperature 25℃; resistance of 100K at the reference temperature; B=3950. The temperature control process of this embodiment is illustrated using the air-cooled heating process of the 3D printer extruder head temperature control system as an example:

[0062] Specifically, such as Figure 2 As shown, the method includes:

[0063] S1: Identify the heating model of the controlled object in the air-cooled heating process, and determine the steady-state gain and time constant in the heating model.

[0064] By identifying the second-order model of the heating process, the following model (transfer function) is obtained:

[0065]

[0066] Where K is the steady-state gain, such as K = 59; T is the time constant, such as T = 61 seconds; ζ is the damping coefficient; y(s) is the temperature output; u(s) is the control input; and s is the frequency domain operator (frequency domain methods are commonly used in control theory).

[0067] The transfer function describes the heating dynamics when u > 0, where there is a significant inertia time between the change in resistance wire voltage and the change in temperature. When u = 0, the temperature decreases due to air cooling while simultaneously being maintained by its own inertia; the cooling dynamics are determined by the system's inertia and the current ambient temperature. During the cooling process, the system is in an open-loop state. To achieve precise temperature control, appropriate constraints need to be placed on the control variables during heating to ensure that the temperature reaches the set point with no or minimal overshoot.

[0068] S2: The heating model is converted into an integral incremental model. The heating state of the heating model is predicted using the integral incremental model to obtain the predicted heating state. The predicted temperature output of the integral incremental model is determined using the integral incremental model and the predicted heating state.

[0069] Based on the actual needs of the air-cooled heating process, G(s) is used to predict future temperature inertial changes using an integral incremental model, thereby further optimizing temperature control. First, G(s) is transformed into a discrete state-space model, with its system matrix being (A... m B m C m ), with state x m (k). Then, to eliminate steady-state error, the discrete model is transformed into an incremental model, forming new state variables x(k) and system matrices (A, B, C). By introducing integration, the steady-state error of the system can be further eliminated, improving the accuracy and stability of temperature control. This method allows the control quantity of the heating process to be reduced in advance, preventing oscillations and overshoot in the system over a period of time.

[0070] Therefore, model (1) is transformed into an integral incremental model, expressed as:

[0071]

[0072] Among them, A m B m C m The system matrix after transforming the heating model G(s) into a discrete state-space model; Δx m (k) represents the change of state variables after transforming the heating model G(s) into a discrete state-space model at time k; A, B, and C are the system matrices after transforming the discrete state-space model into an integral-increment model; x(k) represents the state variables after transforming the discrete state-space model into an integral-increment model; Δu(k) represents the change of control quantity at time k; o m y(k) is a zero matrix of appropriate dimension; y(k) represents the output temperature of the integral incremental model at time k.

[0073] The system sampling period is 0.5 seconds, and the resulting system matrix is:

[0074] C = 0 0 1.

[0075] The current time is k, and the given control time domain is N. c (Calculate future N) c Δu), predict time domain N p (Current state variable x(k) predicts future N) p (N state variables), and generally there are N c <N p Assuming the change in the control quantity outside the control time domain is zero, we have:

[0076] Δu(k+i)=0,i=N c N c +1,…,N p -1

[0077] Using the incremental model described above, the system's state variables can be predicted, and the approximate results for the state variables are as follows:

[0078]

[0079] Where x(k+N) p ) is the Nth prediction of the current heating state variable x(k) p One heating state; N p For the given prediction time domain; N c For a given control time domain, N c <N p The change of the control quantity outside the control time domain is set to zero; Δu(k), Δu(k+1), ... Δu(k+N) c -1) represents the future N that needs to be calculated. c Changes in the control quantity.

[0080] Combining model (2) and state prediction (3), the predicted state of the system is transformed into the predicted output of the system, that is, the expression for the predicted temperature output of the integral incremental model is:

[0081]

[0082] Representing the output in matrix form, the final prediction output of the integral incremental model is:

[0083] Y=Fx(k)+φΔU (4)

[0084] Among them, Y=[y(k+1),y(k+2),…y(k+N p )] T ;

[0085] ΔU=[Δu(k),Δu(k+1),…Δu(k+Nc -1)] T ;

[0086]

[0087] In the formula, y(k+1), y(k+2), ..., y(k+N) p ) represents the predicted N p Temperature output.

[0088] After identifying the heating process model (1), it is transformed into an integral incremental model (2), and the system state prediction (3) and system output prediction (4) are obtained using (2). The inertial change of future temperature is predicted by using an integral incremental model, so that the control quantity can be reduced in advance before the heating process reaches the set temperature. Model predictive control can be optimized online based on real-time measurement results, further improving control accuracy and stability.

[0089] S3: Determine the control quantity constraints and control quantity change constraints of the heating model; the upper limit of the control quantity change in the control quantity change constraint is determined based on the steady-state gain, the time constant, and the upper limit coefficient of the control quantity change.

[0090] The control variable and its variation constraints are designed such that the upper bound of the control variable variation is inversely proportional to the system steady-state gain K and directly proportional to the time constant T. By simultaneously introducing the control variable u and the constraint on its variation Δu, drastic temperature changes caused by sudden changes in the control variable during the heating process are avoided. Since there is no refrigeration device, the cooling process can only rely on controlled cooling and slow automatic heat dissipation. To better control the overshoot, the variation Δu of the control variable is constrained. The upper bound of the control variable variation Δu is... max Inversely proportional to the system's steady-state gain K and directly proportional to the time constant T, satisfying the formula:

[0091] Δu max =k * T / K, (5)

[0092] Where, Δu max k is the upper limit of the control quantity's change. * >0 represents the upper bound coefficient for the change in the control quantity.

[0093] Consider the control constraints as follows:

[0094] u(k)=u(k-1)+Δu(k)≤u max

[0095] Among them, u max It is the maximum control input, determined by the system's physical conditions. When u(k) = 0, the system is in an air-cooled state; when u(k) > 0, it is in a heated state.

[0096] Control time domain Nc The constraints within the vector form are:

[0097] M1ΔU≤N1

[0098] Where M1 and N1 are constraint matrices, and ΔU is N c The Δu vector within.

[0099] Consider the constraint of the change in control quantity Δu:

[0100] Δu(k)≤Δu max

[0101] Control time domain N c The constraints within the vector form are:

[0102] M2ΔU≤N2

[0103] Where M2 and N2 are constraint matrices. Let M = [M1, M2] T N = [N1, N2] T ,have:

[0104] MΔU≤N (6)

[0105] Simultaneously, control quantity u and control quantity change Δu constraints are introduced, where the upper bound of the control quantity change is inversely proportional to the system steady-state gain K and proportional to the upper bound of the time constant T, satisfying equation (5). Since the air cooling process is slower than the heating process, Δu(k) does not need to be bounded, allowing the cooling process to start quickly. The constraint conditions are described in matrix form, resulting in equation (6), which is the final constraint condition.

[0106] Setting an upper limit on the rate of change of the control variable limits its speed of change. Based on the controller's response speed and the system's inertial characteristics, appropriate control parameters are selected to further smooth the heating process and improve the accuracy and stability of temperature control. Preemptively reducing the control variable can prevent excessive fluctuations and overshoot, resulting in a smoother and more stable heating process while improving the accuracy of temperature regulation.

[0107] like Figure 3 As shown, select Δu max =∞, i.e., unconstrained, Δu max =0.5, Δu max =1、Δu max =1.5 A temperature control experiment was conducted. As the control variable changes, the constraints become increasingly relaxed, and the system approaches an unconstrained state. When Δu max When Δu = 1.5, its oscillation and overshoot are completely consistent with the unconstrained case. However, when Δu maxWhen Δu = 0.5, temperature overshoot is almost eliminated, and the oscillation process becomes very small, with the temperature fluctuating slightly below the target value. To balance the steady-state accuracy, oscillation, and overshoot performance of temperature control, Δu is selected as the optimal value. max =1(k * =0.96), which serves as a constraint for the change of control quantity.

[0108] S4: Introduce a control quantity change weight coefficient, and determine the value of the control quantity change weight coefficient and the value of the control quantity change upper bound coefficient based on the overshoot and oscillation degree of the current heating model.

[0109] During the heating process, a control quantity weight matrix (control quantity change weight coefficient) is introduced. upper limit weighting coefficient k of control quantity change * By adjusting these two types of weights, the controller performance can be optimized, large abrupt changes in the control quantity can be overcome, and stable temperature regulation without overshoot or oscillation can be achieved.

[0110] weight matrix Sum of weighting coefficients k * This is a newly introduced normalized adjustable parameter in this scheme. During the implementation of gradual temperature control in the air-cooled heating process, it can be adjusted according to the following rules:

[0111] 1) Weight matrix Sum of weighting coefficients k * Adjustments are made based on a unit matrix or 1.

[0112] 2) If the system oscillates or overshoots significantly, increase the weight matrix. Introduce a larger control cost into the objective function J; conversely, the weight matrix can be appropriately reduced. Speed ​​up system response.

[0113] 3) If the system oscillates or overshoots significantly, decrease k. * This reduces the upper bound of the control variable's change, limiting large jumps in the control variable; conversely, increasing k appropriately reduces the upper bound. * This will speed up the system response.

[0114] S5: Determine the objective function for solving the control quantity change based on the reference temperature input of the heating model, the predicted temperature output of the integral incremental model, and the control quantity change weight coefficient.

[0115] Given an initial system temperature of y0 and a desired temperature of y0. * A smooth temperature reference trajectory can be derived from...

[0116]

[0117] Where, ω c Represents excessive speed, ys The sequence is denoted as Equation (7) achieves the generation of a gradual temperature.

[0118] Based on constraint (6), the ΔU sequence can be determined by solving the following optimization problem.

[0119]

[0120] stMΔU≤N

[0121] This step integrates the calculation results from the preceding steps to update the control variable change Δu(k). In the objective function J, the first term (Y) s -Y) T (Y s -Y) represents the prediction error between the input and output, determined by the reference trajectory Y. s The second term is constructed by subtracting the system prediction output Y obtained in step S2. This is used to prevent excessive control signal actions and serves as a constraint on the control signal. For quadratic programming problems, methods such as the Lagrange multiplier method, the effective set method, the primal-dual method, and the interior point method can all be used for real-time optimization solutions.

[0122] According to the predictive control principle, the first element of ΔU is used for the current control quantity update, Δu(k) = [1 0… 0]ΔU. The control constraint (6) obtained in step S3 constitutes the constraint condition MΔU≤N for the optimization problem. Solve the optimization problem of equation (8), perform real-time constrained optimization for the change of control quantity at each time step, calculate the real-time adjustment value Δu(k) of the change of control quantity, and realize the rolling optimization of the change of control quantity Δu(k) at each time step.

[0123] Pick Set Δu max =1, prediction time domain N p =100, control time domain N c =2. Due to the introduction of control costs, r w The value of r plays a crucial role in suppressing temperature oscillations. An appropriate weight r can be selected experimentally. w .like Figure 4 As shown, r is taken in the experimental test. w =0 (i.e., unweighted), r w =10 and r w = 20, three possibilities.

[0124] After introducing weighting, the overshoot and oscillation of the temperature control curve were significantly reduced. w The larger the value of r, the more cautious the changes in the control variable will be, meaning Δu will be smaller, resulting in a more stable steady state of the temperature process; on the other hand, r wA larger value will also slow down the temperature response speed of the heating process, mainly manifested in the dynamic process of temperature change. Therefore, increasing the weight of the control constraint is a control method that overcomes overshoot and oscillations at the cost of system response speed. This method is highly practical for steady-state air-cooled heating processes. Therefore, r is selected... w =20 is used as the controller parameter.

[0125] To verify the effectiveness of the method, a temperature regulation experiment was conducted using the designed constrained temperature predictive controller, with r w =20, Δu max =1, with the target temperature set at 200℃. Simultaneously, under the condition of consistent response speed, a PID controller is designed for comparison. For example... Figure 5 As shown, even when the response speeds of the two types of controllers are the same, the overshoot of PID control is significantly greater than that of the method proposed in this invention. This is because PID control can only react after a deviation occurs. When the temperature approaches the target value, although the derivative action can produce a certain deceleration effect, the lack of a cooling process prevents PID from producing an accurate reverse regulation effect, causing the temperature to continue to increase. Furthermore, conventional PID controllers fail to incorporate the dynamic characteristics of the system and lack a constraint mechanism for the control quantity, thus failing to effectively overcome the overshoot and oscillation of large inertia systems. In contrast, the design method proposed in this invention fully considers the heating characteristics under air-cooled regulation and specifically introduces control quantity constraints to ensure that the temperature enters a steady state as smoothly as possible.

[0126] S6: Solve the current control quantity change based on the objective function of the control quantity change and the control quantity constraints and control quantity change constraints, determine the current control quantity of the heating model based on the current control quantity change, and perform temperature control on the heating model according to the current control quantity.

[0127] In this embodiment, ① by identifying a second-order model of the heating process and predicting the inertial change of future temperature using an incremental model, the control quantity can be reduced in advance before the set temperature is reached, thus controlling the heating process more precisely. ② By introducing constraints on the control quantity and its changes, drastic temperature changes caused by sudden changes in the control quantity are avoided during the heating process, thereby ensuring the stability and accuracy of the control process. Incremental models, control quantity constraints, and control quantity change constraints are incorporated into the controller design to achieve precise control of the heating process. ③ Control quantity weighting coefficients and control quantity change upper bound coefficients are introduced. By adjusting these two types of coefficients, the temperature of the heating process can be controlled more precisely.

[0128] Example 2

[0129] This embodiment provides a model predictive temperature control system for an air-cooled heating process, the system comprising:

[0130] The heating model identification module is used to identify the heating model of the controlled object in the air-cooled heating process and determine the steady-state gain and time constant in the heating model.

[0131] The model prediction module is used to convert the heating model into an integral incremental model, use the integral incremental model to predict the heating state of the heating model, obtain the predicted heating state, and use the integral incremental model and the predicted heating state to determine the predicted temperature output of the integral incremental model.

[0132] The constraint construction module is used to determine the control quantity constraints and control quantity change constraints of the heating model; the upper limit of the control quantity change in the control quantity change constraint is determined based on the steady-state gain, the time constant, and the upper limit coefficient of the control quantity change.

[0133] The weight coefficient determination module is used to introduce the control quantity change weight coefficient and determine the value of the control quantity change weight coefficient and the value of the control quantity change upper bound coefficient based on the overshoot and oscillation degree of the current heating model.

[0134] The objective function construction module is used to determine the objective function for solving the control quantity change based on the reference temperature input of the heating model, the predicted temperature output of the integral incremental model, and the control quantity change weight coefficient.

[0135] The temperature control module is used to solve the current control quantity change based on the objective function of the control quantity change and the control quantity constraints and control quantity change constraints, determine the current control quantity of the heating model based on the current control quantity change, and perform temperature control on the heating model according to the current control quantity.

[0136] Example 3

[0137] This embodiment provides an electronic device, including a memory and a processor. The memory is used to store computer programs, and the processor runs the computer programs to enable the electronic device to perform the model prediction temperature control method for the air-cooled heating process of Embodiment 1.

[0138] Alternatively, the aforementioned electronic device may be a server.

[0139] In addition, embodiments of the present invention also provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the model-predictive temperature control method for the air-cooled heating process of Embodiment 1.

[0140] Embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0141] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0142] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0143] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0144] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0145] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A model predictive temperature control method for an air-cooled heating process, characterized by, The method comprises: identifying a heating model of a controlled object in an air cooling heating process, and determining a steady-state gain and a time constant in the heating model; converting the heating model into an integral type incremental model, predicting a heating state of the heating model by using the integral type incremental model to obtain a predicted heating state, and determining a predicted temperature output of the integral type incremental model by using the integral type incremental model and the predicted heating state; determining a control quantity constraint and a control quantity change constraint of the heating model; an upper limit value of the control quantity change in the control quantity change constraint is determined according to the steady-state gain, the time constant and a control quantity change upper limit coefficient; introducing a control quantity change weight coefficient, and determining a value of the control quantity change weight coefficient and a value of the control quantity change upper limit coefficient according to an overshoot degree and an oscillation degree of the current heating model; determining a solving target function of the control quantity change according to a reference temperature input of the heating model, the predicted temperature output of the integral type incremental model and the control quantity change weight coefficient; solving the current control quantity change based on the solving target function of the control quantity change and the control quantity constraint and the control quantity change constraint, determining a current control quantity of the heating model based on the current control quantity change, and performing temperature control on the heating model according to the current control quantity; wherein the calculation formula of the upper limit value of the control quantity change in the control quantity change constraint is: Δu max = k * T / K where Δu max is the upper limit value of the control amount variation; k * > 0 is the control amount variation upper limit coefficient; K is the steady-state gain; and T is the time constant.

2. The method of claim 1, wherein, the expression of the heating model is: wherein K is a steady-state gain; T is a time constant; ζ is a damping coefficient; y(s) is a temperature output; u(s) is a control input; and s is a frequency domain operator.

3. The method of claim 2, wherein, the expression of the integral type incremental model is: wherein A m , B m , C m are system matrices after the heating model G(s) is converted into a discrete state space model; Δx m (k) is a change in a state variable at time k after the heating model G(s) is converted into a discrete state space model; A, B, C are system matrices after the discrete state space model is converted into an integral incremental model; x(k) is a state variable after the discrete state space model is converted into an integral incremental model; Δu(k) represents a change in a control variable at time k; O m is a zero matrix; and y(k) represents an output temperature of the integral incremental model at time k.

4. The method of claim 3, wherein, the prediction expression of the predicted heating state is: where x(k + N p ) is the N p th predicted heating state of the current heating state variable x(k); N p is a given prediction time domain; N c is a given control time domain, N c <N p , the control variable change outside the control time domain is set to zero; Δu(k), Δu(k + 1), … Δu(k + N c -1) are the future N c control variable changes to be calculated.

5. The method of claim 4, wherein, the expression of the predicted temperature output of the integral type incremental model is: Y = Fx(k) + φΔU where Y = [y(k+1), y(k+2),... y(k+N p )] T ; ΔU = [Δu(k), Δu(k+1),... Δu(k+N c -1)] T ; where y(k+1), y(k+2),... y(k+N) represent the predicted N p temperature outputs. p ​ 6. The method of claim 1, wherein, The value of the control quantity change weight coefficient and the value of the control quantity change upper limit coefficient are determined according to the overshoot degree and the oscillation degree of the current heating model, and specifically comprising: the control quantity change weight coefficient and the control quantity change upper limit coefficient are adjusted based on a unit matrix or a preset value; when the overshoot degree is greater than a first preset value or the oscillation degree is greater than a second preset value, the current control quantity change weight coefficient is increased and the current control quantity change upper limit coefficient is decreased; otherwise, the current control quantity change weight coefficient is decreased and the current control quantity change upper limit coefficient is increased.

7. The method of claim 5, wherein, the calculation formula of the reference temperature input of the heating model is: where y s is the reference temperature input; y0 is the initial temperature; y * is the desired temperature; ω c represents the over- or under-speed.

8. The method of claim 5, wherein, the expression of the solving target function of the control quantity change is: wherein, is the control variation weight coefficient matrix; the first element of the solved ΔU is used for the current control quantity update, Δu(k) = [1 0…0]ΔU; y s The generated temperature sequence is denoted as 9. A model predictive temperature control system for an air-cooled heating process, characterized by, The system comprises: a heating model identification module, configured to identify a heating model of a controlled object in an air cooling heating process, and determine a steady-state gain and a time constant in the heating model; a model prediction module, configured to convert the heating model into an integral type incremental model, predict a heating state of the heating model by using the integral type incremental model to obtain a predicted heating state, and determine a predicted temperature output of the integral type incremental model by using the integral type incremental model and the predicted heating state; A constraint building module is configured to determine a control quantity constraint and a control quantity change constraint of the heating model; an upper limit value of the control quantity change in the control quantity change constraint is determined according to the steady-state gain, the time constant and a control quantity change upper limit coefficient; wherein the calculation formula of the upper limit value of the control quantity change in the control quantity change constraint is: Δu max = k * T / K where Δu max is the upper limit of the control variable change; k * > 0 is the upper limit coefficient of the control variable change; K is the steady-state gain; and T is the time constant. A weight coefficient determining module is configured to introduce a control quantity change weight coefficient, and determine the value of the control quantity change weight coefficient and the value of the control quantity change upper limit coefficient according to the overshoot degree and the oscillation degree of the current heating model; A target function building module is configured to determine a solving target function of the control quantity change according to the reference temperature input of the heating model, the predicted temperature output of the integral type increment model and the control quantity change weight coefficient; A temperature control module is configured to solve the current control quantity change based on the solving target function of the control quantity change and the control quantity constraint and the control quantity change constraint, determine the current control quantity of the heating model based on the current control quantity change, and perform temperature control on the heating model according to the current control quantity.

Images