Energy self-coordination control method for pump-controlled hydraulic cushion energy storage system
By designing an energy self-coordination control method in the pump-controlled hydraulic pad system, using sliding mode prediction algorithm and related components, the problem of instability in energy storage and reuse in traditional systems is solved, efficient and stable energy management is achieved, and system performance and equipment stability are improved.
Patent Information
- Application Number
- CN202510189844.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-20
- Publication Date
- 2025-06-20
AI Technical Summary
Traditional pump-controlled hydraulic pad systems have challenges in the effective storage and reuse of energy, resulting in unstable energy recovery, which may cause system oscillation and affect the accuracy and production efficiency of injection molded parts.
A pump-controlled hydraulic pad energy storage system energy self-coordinated control method is designed, and an energy self-coordinated control strategy is adopted, including energy storage mode, standby mode and energy supply mode. Accurate energy control is carried out through sliding mode prediction algorithms, and components such as supercapacitors, safety resistors and motor drives are used to ensure stable energy control.
It realizes efficient energy recovery and storage, and achieves precise and stable control during the energy release process, improves the overall performance of the system, reduces energy consumption and operation and maintenance costs, and enhances the stable operation and life of the equipment.
Smart Images
Figure CN120185018A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of energy storage system control, and particularly relates to an energy self-coordination control method for a pump-controlled hydraulic pad energy storage system. Background Art
[0002] The pump-controlled hydraulic pad system is widely used and penetrates into many important industries such as industrial manufacturing, heavy machinery, aerospace, etc., and has high value for improving equipment efficiency and reducing energy consumption. However, in the actual operation process, the system faces many challenges, especially in the effective storage and reuse of energy. These problems directly restrict the further improvement of the overall performance of the system. Taking an injection molding machine in industrial manufacturing as an example, the pump-controlled hydraulic pad system, as one of its core components, is responsible for providing stable pressure and speed control during the injection molding process. When dealing with the regenerative electric energy generated by the system, traditional control methods often lead to unstable energy recovery due to imperfect control strategies, and may even cause system oscillation, affecting the accuracy and production efficiency of injection molded parts. Specifically, when the hydraulic pump generates excess pressure energy during operation, if this part of the energy cannot be effectively recovered and stored, it will not only cause energy waste, but also may impact the power grid due to improper energy feedback, increasing additional energy consumption and operation and maintenance costs.
[0003] In addition, during the energy release stage, traditional control methods often have difficulty achieving precise regulation and smooth transition of energy. For example, in working conditions that require rapid response, such as emergency braking or rapid startup, the system may experience unstable energy release due to control lag or overshoot, thereby affecting the stable operation and lifespan of the equipment. Especially in the field of high-precision machining, this instability may lead to a decline in product quality and even cause production accidents.
[0004] Therefore, how to design a strategy that can not only efficiently recover and store regenerative electric energy, but also achieve precise and stable control during the energy release process has become an urgent need for the technical development of the pump-controlled hydraulic pad system. The present invention proposes an energy self-coordination control method for a pump-controlled hydraulic pad energy storage system, which is an innovative solution proposed specifically for this technical problem, aiming to achieve efficient utilization and smooth transition of energy through advanced control strategies, thereby promoting the overall upgrade and application expansion of the pump-controlled hydraulic pad system technology. Summary of the Invention
[0005] The present invention designs a pump-controlled hydraulic cushion energy storage system according to the idea of self-production and self-use, and proposes an energy self-coordination control strategy for the energy storage system to improve the energy utilization efficiency. The pump-controlled hydraulic cushion energy storage system adopts an energy self-coordination control strategy to ensure the high efficiency of energy utilization, including an energy storage mode, a standby mode, and a power supply mode, and realizes flexible conversion through mode switching judgment. Under multiple condition constraints, the system operates stably, uses a sliding mode prediction algorithm for precise energy control, and uses components such as super capacitors, safety resistors, and motor drives to jointly act on the stable control of energy to ensure the safety and high efficiency of the system.
[0006] The present invention provides an energy self-coordination control method for a pump-controlled hydraulic cushion energy storage system. The energy storage system has three modes: an energy storage mode, a standby mode, and a power supply mode. The energy self-coordination control method for the energy storage system includes the following processes: S1. Design an energy self-coordination control strategy for the pump-controlled hydraulic cushion energy storage system. The energy self-coordination control strategy includes multi-condition switching of energy storage or power supply logic and stable input or output control of energy. S2. Multi-condition switching of energy storage or power supply logic: Switch the working mode of the energy storage system according to the change of the DC bus voltage of the energy storage system, the die pad displacement of the pump-controlled hydraulic cushion, and the change of the super capacitor of the energy storage system. S3. Stable output or input control of energy: The energy storage mode is used for the management of the inductor current and capacitor voltage on the energy storage side of the bidirectional DC / DC converter, and the power supply mode is used for the regulation of the inductor current and bus voltage on the bus side of the bidirectional DC / DC converter. Both the energy storage mode and the power supply mode achieve precise coordination through a sliding mode predictive control algorithm.
[0007] Further, the sliding mode predictive control algorithm includes: S31. Establish the state space equations of the bidirectional DC / DC converter in the energy storage mode and the power supply mode. S32. Construct a sliding mode surface according to the desired output / input voltage and inductor current. S33. Determine the sliding mode switching function according to the sliding mode surface, and combine the feedback correction and rolling optimization ideas of model predictive control to achieve the optimal control of the switching function. S34. Select an appropriate sliding mode reaching law to ensure that the system quickly and stably reaches the sliding mode surface. S35. Select and substitute for calculation according to the algorithm formula, and finally solve the expression of the controller. S36. Verify the stability of the system through the Lyapunov function.
[0008] Further, the energy storage mode, the standby mode, and the power supply mode cover four working stages of the pump-controlled hydraulic cushion. The four working stages respectively include the die pad descending in the pre-acceleration stage, the die pad being pushed down in the stretching stage, the stage of maintaining the blank holding force, and the stage of the die pad ascending and returning.
[0009] Further, during the blank holding force stage, the die cushion upward return stroke stage, the pre-acceleration stage, and the stretching stage, the energy storage system monitors the DC bus voltage and selects to enter the standby mode, the energy storage mode, or the power supply mode according to the DC bus voltage. If the voltage is less than the specified range, the power supply mode is started. If it is greater than the specified range, the energy storage mode is started; if the voltage is within the specified range, the standby mode is entered; when the charge value is not between 0 and the maximum value, the safety system is entered, and the braking resistor protection in the safety system is started. The regenerative energy in the working cycle is consumed by the braking resistor.
[0010] Further, the implementation steps of the sliding mode predictive controller of the energy storage system in the energy storage mode include: Convert the static voltage and current into a dynamic equation related to time and define S The switching function of 1 is as follows: ; Where: T is the switching period, D 1 is the duty cycle; According to Kirchhoff's law, the state equation of the energy storage system in the energy storage mode is: ; In the formula, is the voltage of the supercapacitor bank; is the voltage at the load end; i L is the inductor current; is the equivalent resistance of the supercapacitor; is the inductor; The established space state equation of the bidirectional DC / DC converter in the energy storage mode is sorted into a matrix equation as: ; Adopt sliding mode predictive control for the bidirectional DC / DC converter of the supercapacitor energy storage system. Combining the required controlled capacitor charging voltage U sc and the inductor current i L as the basis, take the capacitor charging voltage error e 1( t ) as: ; In the formula, U scr is the expected capacitor charging voltage; The rate of change of the capacitor voltage error e 2( t ) is: ; For the above e 1( t ), e 2( t ) take the derivative and organize to obtain the following formula: ; Organize it into the state - space equation to get: ; Where: , , , ; Construct the sliding surface of the sliding - mode prediction model s 1( t ) = 0: ; In the formula, a 1, a 2 are the sliding coefficients of the sliding surface; According to the sliding surface, construct the sliding - mode switching function, and define the sliding - mode switching function s 1( t ) as: ; In the formula, a 1 a 2 > 0, is the sliding - coefficient matrix of the sliding surface; E 1( t ) is the sliding - mode variable matrix; Take the derivative of the sliding - mode switching function to obtain s 1 ´ ( t ): ; Define the sign function sgn s 1( t ) as: ; The form of the equal - speed reaching law is as follows: ; 1 represents the rate when approaching the sliding surface; The form of the power - reaching law is as follows: ; The form of the exponential - reaching law is as follows: ; The form of the general reaching law is as follows: ; Thus, it is obtained that: ; Furthermore, it is obtained that u 1 expression: ; By adopting voltage outer loop control and current inner loop control, it is known from that: ; The expression of the controller U is obtained as: ; Select the Lyapunov function L 1 to verify the stability of the system as follows: ; Taking the derivative, it is obtained that: ; According to the sliding mode switching function s ( t ), the s ( t + 1) sliding mode prediction model is obtained: ; According to the sliding mode prediction model, the sliding mode surface, that is, the prediction value of the sliding mode switching function s ( t + p ) at a certain moment is as follows: ; A feedback correction link is introduced into the system to correct and adjust the predicted value of the sliding mode switching function at a future moment. The predicted value of t - p at the moment for t is s ( t | t - p ): ; The predicted value after feedback correction is as follows: ; Add the correction coefficient h p ∈ R . As h p increases, the role of feedback correction will increase, and vice versa; According to the predictive control theory, the original sliding mode switching function is changed: ; n is the system approach speed parameter, n > 0, s r ( t + p ) is the reference value of the sliding mode switching surface at the future t + p moment, T s is the system sampling period, nT s > 0; When u 1 = 0: ; The expression of the controller changes to: ; When u 1 = 1: ; The expression of the controller changes to: ; After sorting, the final control function is obtained: ; Based on the sliding mode predictive control theory, a control function is established to achieve stable control of the energy of the bidirectional DC / DC converter in the energy storage mode.
[0011] Furthermore, the implementation steps of the sliding mode predictive controller of the energy storage system in the power supply mode include: Convert the static voltage and current into a dynamic process related to time. The supercapacitor U sc terminal provides electrical energy, and the common DC bus side is the load. S The switching function of 2 is as follows: ; Where: R is the load terminal resistance, D 2 is the duty cycle; According to Kirchhoff's law, the state equation of the energy storage system in the power supply mode is: ; The established space state equation of the bidirectional DC / DC converter in the power supply mode is sorted into a matrix equation: ; Adopt sliding mode predictive control for the bidirectional DC / DC converter of the supercapacitor energy storage system, combined with the bus expected voltageU dc and the inductor current i L Based on this, the bus voltage error e 3( t ) is defined as: ; In the formula, U dcr is the expected bus voltage; The rate of change of the bus voltage error e 4( t ) is defined as: ; For the above e 1( t ), e 2( t ) are differentiated and sorted out to obtain the following formula: ; Sorted into the state space equation to get: ; Among them: , , ; Construct the sliding surface of the sliding mode prediction model s 2( t ) = 0: ; In the formula, a 3, a 4 are the sliding coefficients of the sliding surface; According to the sliding surface, construct the sliding mode switching function, and define the sliding mode switching function s 2( t ) as: ; In the formula, a 3 a 4 > 0, A 2 is the sliding coefficient matrix of the sliding surface, E 2( t ) is the sliding mode variable matrix; Derive the sliding mode switching function to get s 2 ´ ( t ): ; ; Define the sign function sgn s 2(t ): ; After the reaching law is selected, we get: ; Furthermore, we get u 2 expressions: ; By adopting voltage outer loop control and current inner loop control, we know that: ; After integrating both sides and simplifying, we get the expression of the controller U : ; Select the Lyapunov function to verify the stability of the system as follows: ; Taking the derivative, we get: ; According to the sliding mode switching function s ( t ), we get s ( t + 1) sliding mode prediction model: ; According to the sliding mode prediction model, the sliding mode surface, that is, the predicted value of the sliding mode switching function s ( t + p ) at a certain moment is as follows: ; Introduce a feedback correction link in the energy storage system to correct and adjust the predicted value of the sliding mode switching function at a future moment. The predicted value of t - p at the moment of t for s ( t | t - p ) is: ; The predicted value after feedback correction is as follows: ; Add the correction coefficient h p ∈ R . As h p increases, the role of feedback correction will increase, and vice versa; According to the predictive control theory, the original sliding mode switching function is changed as follows: ; n is the system approaching speed parameter, n > 0, s r ( t + p ) is the reference value of the sliding mode switching surface at the future t + p moment, T s is the system sampling period, nT s > 0; The predictive model of the sliding mode switching function is obtained: When u 2 = 0: ; The expression of the controller changes to: ; When u 2 = 1: ; The expression of the controller changes to: ; After sorting, the final control function is obtained: ; Based on the sliding mode predictive control theory, a control function is established to achieve stable control of the energy of the bidirectional DC / DC converter in the power supply mode.
[0012] The beneficial effects of the present invention compared with the prior art are as follows: (1) Based on the idea of self-production and self-use, the present invention designs an energy self-coordination control method for a pump-controlled hydraulic pad energy storage system, and proposes an energy self-coordination control strategy for the energy storage system to improve the energy utilization efficiency and high efficiency of energy utilization; (2) The present invention includes an energy storage mode, a standby mode, and a power supply mode, and realizes flexible conversion through mode switching judgment. Under multiple condition constraints, the energy storage system can still operate stably; (3) The present invention uses a sliding mode prediction algorithm for precise energy control, and components such as supercapacitors, safety resistors, and motor drives act together on the stable control of energy, thus ensuring the safety and high efficiency of the energy storage system; (4) The present invention gives the working process of the energy storage system when the pump-controlled hydraulic pad is in a power generation, no-power generation, or no-power generation standby state. By combining sliding mode control and model state prediction control, precise and stable control of the energy of the pump-controlled hydraulic pad energy storage system is achieved; (5) The present invention enhances the adaptability of the control algorithm by using the sliding mode prediction algorithm: by predicting the sliding mode switching function, the change of the system state can be predicted within a certain period of time in the future, so as to better adapt to the dynamic change of the system, which helps the controller to respond to the system change more quickly and adopt appropriate control strategies; (6) The present invention reduces the jitter in the control process by using the sliding mode prediction algorithm: the jitter of the control system refers to the oscillation that may occur when the system reaches the desired state. By predicting the sliding mode switching function, the control input can be adjusted more precisely, reducing the system jitter, thereby improving the smoothness and stability of the control; (7) The present invention optimizes the performance of the control algorithm by using the sliding mode prediction algorithm: the predicted sliding mode switching function can be used to optimize the generation of the control input to achieve better control performance. By considering the change of the future state, the controller can better plan the control strategy to meet the performance requirements. BRIEF DESCRIPTION OF THE DRAWINGS
[0013] Figure 1 It is the flowchart of the mode self-selection control of the energy storage system of the present invention.
[0014] Figure 2 It is the schematic diagram of the sliding mode control process of the present invention Figure 1 。
[0015] Figure 3 It is the schematic diagram of the sliding mode control process of the present invention Figure 2 。
[0016] Figure 4 It is the basic schematic diagram of the model state prediction control of the present invention.
[0017] Figure 5 It is the flowchart of selecting the optimal switch state in the energy storage mode of the present invention.
[0018] Figure 6 It is the flowchart of selecting the optimal switch state in the power supply mode of the present invention. Specific Embodiments
[0019] The present invention will be further described below in conjunction with specific embodiments. The illustrative embodiments and explanations of this invention are used to explain the present invention, but do not limit the present invention.
[0020] Embodiment: As Figures 1 - 6 shown, the present invention proposes an energy self - coordination control method for a pump - controlled hydraulic pad energy storage system, including the following processes: Design an energy self - coordination control strategy for the pump - controlled hydraulic pad energy storage system. The energy self - coordination control strategy includes multi - condition switching of energy storage or energy supply logic and stable input or output control of energy.
[0021] Multi - condition switching of energy storage or energy supply logic: According to the change of the DC bus voltage of the energy storage system, the die pad displacement of the pump - controlled hydraulic pad, and the change of the supercapacitor ( SOC ) of the energy storage system, switch the working mode of the energy storage system. The switching process is as follows: Calculate the charge number ( SOC ) value of the supercapacitor. When the charge number ( SOC ) value is between 0 and the maximum value, collect the die pad displacement data of the pump - controlled hydraulic pad. According to the charge number ( SOC ) of the supercapacitor and the die pad displacement of the pump - controlled hydraulic pad, judge the current working stage. When the charge number ( SOC ) is out of range, the regenerative energy in this working cycle is consumed by the braking resistor; when the charge number ( SOC ) limit is within the normal range, continue to collect the displacement of the die pad, and compare it with the expected hydraulic pad displacement curve. Judge the current working stage of the hydraulic pad through the die pad displacement and time, and match different modes. The energy storage system includes three modes: namely, energy storage mode, standby mode, and energy supply mode.
[0022] The energy storage mode, standby mode, and energy supply mode cover four working stages of the pump - controlled hydraulic pad, including the die pad descending in the pre - acceleration stage, the die pad being pushed down in the stretching stage, the stage of maintaining the blank - holding force, and the stage of the die pad ascending and returning.
[0023] In the stage of maintaining the blank - holding force, the stage of the die pad ascending and returning, the pre - acceleration stage, and the stretching stage, the energy storage system monitors the DC bus voltage and selects to enter the standby mode, energy storage mode, or energy supply mode according to the DC bus voltage. If the voltage is less than the specified range, start the energy supply mode (Boost boost mode); if it is greater than the specified range, start the energy storage mode (Buck buck mode); if the voltage is within the specified range, enter the standby mode; when the charge number ( SOC ) value is not between 0 and the maximum value, enter the safety system, start the braking resistor protection in the safety system, and the regenerative energy in this working cycle is consumed by the braking resistor.
[0024] Monitor the common DC bus voltage of the motor in different working stages. Since there is regenerative electric energy during the stretching stage, after it accumulates in the bus, it will cause the bus voltage to rise; after the energy is stored, the voltage can drop to the normal value; when the motor outputs load force to do work, the DC bus of the motor will have a drop phenomenon; based on this, the energy storage system mode can be coordinated, and the stored energy can be supplied to the motor for use. Thus, the multi-condition logic mode self-selection is completed. Subsequently, when the energy storage system boosts voltage to supply energy or steps down voltage for energy storage, it is necessary to control the bidirectional DC / DC converter to output stable voltage and current.
[0025] The charge number of the supercapacitor ( SOC ), and the calculation process is as follows: Calculate the maximum energy stored in the supercapacitor bank: (1) In the formula: is the maximum energy stored in the supercapacitor bank (J), is the capacitance of the supercapacitor bank (F), is the upper limit of the terminal voltage of the supercapacitor bank (V), is the lower limit of the terminal voltage of the supercapacitor bank (V).
[0027] Calculate the charge number of the supercapacitor ( SOC) ): (2) In the formula: E cmax is the upper limit of the maximum energy storage of the supercapacitor (J), is the voltage of the supercapacitor bank (V); Stable output or input control of energy: The energy storage mode is used for the management of the inductor current and capacitor voltage on the energy storage side of the bidirectional DC / DC converter, and the energy supply mode is used for the regulation of the inductor current and bus voltage on the bus side of the bidirectional DC / DC converter; both the energy storage mode and the energy supply mode are precisely coordinated through the sliding mode predictive control algorithm.
[0029] The sliding mode predictive control algorithm includes sliding mode control and model state predictive control. The sliding mode predictive control uses the sliding mode controller as the main body, applies the ideas of feedback correction and rolling optimization in the model state predictive control, establishes the target evaluation function to obtain the optimal switching solution of the sliding mode controller, and finally realizes the control target of the sliding mode predictive control algorithm of the energy storage system for stable input or output of energy.
[0030] Sliding mode control is a control method based on a sliding mode surface. Its control idea is to establish a sliding mode surface and enable the system to slide to this surface quickly and stably. The system state can slide rapidly on the sliding mode surface and finally achieve the desired control effect. Sliding mode control is a variable structure control within the category of nonlinear control strategies. It switches the state of the space state function to guide the sliding behavior of the system on the sliding mode surface.
[0031] Sliding mode variable structure control can be divided into several main processes: First, the control variable of the system to be controlled needs to reach the set sliding mode surface. This process of sliding change is the reaching stage of sliding mode variable structure control. Then, after the control variable reaches the sliding mode surface and moves towards the equilibrium point on this surface, this process is called the sliding stage of sliding mode variable structure control. Finally, the control quantity moves on the sliding mode surface and stays at the desired stable point, with fluctuations within the limit range.
[0032] The design of the sliding mode control algorithm needs to consider the following three conditions: First, the sliding mode exists; second, the reachability condition is satisfied: any state point outside the switching surface s ( x ) = 0 will reach the switching surface within a finite time; finally, the stability of the sliding mode motion and the dynamic quality requirements of the system are achieved. Each stage of the system's motion is controlled by the switching function s ( x ) and the control law u . Therefore, the motion quality of the system can be improved by selecting different switching functions and control laws.
[0033] Since the bidirectional DC / DC converter in the energy storage system operates in a time-varying, nonlinear, and periodic manner during operation, it belongs to the category of variable structure control systems. Therefore, this provides a theoretical basis for the application of sliding mode control in bidirectional DC / DC converters. However, during the implementation of the sliding mode control algorithm, due to the high-frequency switching of the switches, chattering problems occur, which cannot be ignored in the actual process. The following is a brief explanation of the generation of chattering: When the control system of the bidirectional DC / DC converter moves near the switching surface, the time lag of the switching tube causes a time delay for the controlled system to cross the switching surface. The amplitude of the control quantity decreases as the state quantity decreases, and a decaying triangular wave will appear on the sliding mode surface. The spatial lag of the switch will cause a dead zone for the state quantity in the state space, and an equal-amplitude triangular wave will appear on the sliding mode surface. When the system has not reached the sliding mode surface, the control function switches, that is, the switching action occurs on the surface of the cone with the equilibrium point as the origin. The larger the cone, the stronger the chattering.
[0034] Model predictive control is one of the advanced control algorithms. This method considers the nonlinear model of the system and combines the existing input conditions to predict the output state of the controlled variable within a certain period in the future. By solving the multi-condition constrained optimal state control problem, a smaller tracking error of the system within a certain period in the future can be obtained, which is also one of the advantages of model state prediction. The comparison of the trajectories of the system state changing with time includes the desired trajectory and the actual trajectory. At the initial moment k , the system is in the "previous state", and then under the action of the control input u ( k ), u ( k + 1), u ( k + 2), etc., the system state gradually evolves and tends to the future stable state. The two trajectories intersect at a certain point, and the desired trajectory is generally above the actual trajectory, clearly showing the difference and the intersection point between the two. At the same time, the marked "100" may represent a key data point or a threshold. The basic principle of the control of model state prediction u ( k ), y ( k ) are the control quantity and the output quantity at the moment of k respectively. Within the future finite time domain k + p , the predicted output quantity is minimized with the variance of the reference trajectory to obtain the optimal control quantity at this moment, and the above behavior is repeated.
[0035] Apply the ideas of feedback correction and rolling optimization in the model state prediction algorithm to the sliding mode controller to obtain the optimal solution of the switching state in the next process of the converter. Through this improved method, the chattering problem caused by the high-frequency switching state in the sliding mode control is solved, and at the same time, the robustness of the system is increased. The sliding mode prediction control is supplemented in the bidirectional DC / DC converter energy control system to ensure the control quality. The implementation process of the "sliding mode prediction control algorithm" composed of the sliding mode variable structure and the model state prediction is as follows: (1) Establish the state space equations in the energy storage mode and the power supply mode of the bidirectional DC / DC energy converter; (2) Select the sliding mode surface, and the construction of the sliding mode surface is based on the desired set value. In this control scheme, the output / input voltage of the converter, the inductor current, and the corresponding expected values are used to construct the tracking error to construct the sliding mode surface; (3) Determine the sliding mode switching function according to the sliding mode surface, and apply the ideas of rolling optimization and feedback correction of the model predictive control to the switching state selection of the sliding mode switching function to achieve the optimal control of the switching function; (4) Select an appropriate sliding mode reaching law to ensure that when the control quantity converges rapidly during the reaching stage, it reaches the sliding mode at a relatively fast speed and is finally controlled within the transformation constraint threshold of the expected output value; (5) Select and substitute for calculation according to the algorithm formula, and finally solve the expression of the controller; (6) Select a Lyapunov function to verify that the system satisfies reaching, existence, stability, and has good dynamic quality. Based on this, the sliding mode control can function properly.
[0036] The specific implementation process of the sliding mode predictive control algorithm is as follows: Taking the sliding mode control architecture as the main body, complete the control of the switching tubes in the bidirectional DC / DC converter; To improve the chattering problem of the sliding mode control within the target threshold range, apply the sliding mode switching function, the idea of rolling optimization in model state prediction, and feedback correction to the algorithm, collect key state information of the system, and calculate the switching states of the switching tubes 1 and 2 in the bidirectional converter according to the sliding mode prediction model respectively. When the switching states are 1 and 0, the predicted values of the switching function at the future + moment are obtained; Then compare the magnitudes of the output power and the load power to determine the working mode of the bidirectional converter at this time. Compare the magnitudes of the two objective function values under different working modes, and use the switching state corresponding to the minimum objective function value as the switching tube state at the next moment to control the action of the converter switching tubes. S 1 and S 2 when the switching states are 1 and 0, the predicted values of the switching function at the future k + p moment; Then compare the magnitudes of the output power and the load power to determine the working mode of the bidirectional converter at this time. Compare the magnitudes of the two objective function values under different working modes, and use the switching state corresponding to the minimum objective function value as the switching tube state at the next moment to control the action of the converter switching tubes.
[0037] The advantages of using the sliding mode predictive control algorithm are as follows: (1) Enhance the adaptability of the control algorithm: By predicting the sliding mode switching function, the changes in the system state can be predicted within a certain period in the future, thus better adapting to the dynamic changes of the system. This helps the controller respond more quickly to system changes and adopt appropriate control strategies; (2) Reduce the jitter in the control process: The jitter in the control system refers to the oscillation that may occur when the system reaches the desired state. By predicting the sliding mode switching function, the control input can be adjusted more precisely, reducing system jitter, thereby improving the smoothness and stability of the control; (3) Optimize the performance of the control algorithm: The predicted sliding mode switching function can be used to optimize the generation of the control input to achieve better control performance. By considering the changes in future states, the controller can better plan the control strategy to meet the performance requirements.
[0038] The purpose of the predicted sliding mode switching function in the sliding mode predictive control is to enhance the adaptability of the control, reduce jitter, and optimize the control performance. By introducing the prediction of future states, the controller can adjust the control input more accurately, thereby achieving better control effects.
[0039] The sliding film predictive control uses the model predictive controller and the sliding mode controller to work together to calculate the state of the switch tube, predict the switching function, and correct it through model output feedback, and finally output an accurate switching signal to achieve efficient and intelligent management of the entire energy storage system; according to the S2 logic multi-condition switching control block diagram, the system starts to calculate the state parameters of the energy storage system and obtain the current storable energy value of the energy storage system.
[0040] The control function is established based on the sliding mode predictive control theory to achieve stable energy control in the energy storage mode of the bidirectional DC / DC energy converter, that is, the energy control algorithm is expressed as stable output control of the voltage in the buck output mode. The supercapacitor mode is constant power charging, that is, the voltage of the capacitor will gradually increase, and the energy will be continuously stored. At the same time, the supercapacitor current will decrease to ensure constant power, that is, matching the constant power of power generation.
[0041] Design of sliding mode predictive controller for energy storage system in energy storage mode: (1) Without considering the saturation characteristics of the inductor and the nonlinear characteristics of the output capacitor, the static voltage and current are converted into a time-dependent dynamic process. T is the switching cycle, D 1 is the duty cycle, defined as S The switch function of 1 is as follows: (3) According to Kirchhoff's law, the state equations of the system in the two pressure-reducing modes are: (4) In the formula, is the voltage of the supercapacitor bank (V); is the load terminal voltage (V); i L is the inductor current (A); is the equivalent resistance of the supercapacitor (Ω); is the inductance (H).
[0044] The spatial state equation of the bidirectional DC / DC converter in the energy storage mode established by combining the two equations is organized into a matrix equation: (5) (2) The bidirectional DC / DC converter of the supercapacitor energy storage system is controlled by sliding mode predictive control, combined with the capacitor charging voltage to be controlled. U sc (For constant current charging, an equivalent resistor can be connected in series) and the inductor current i L Based on the capacitor charging voltage error e 1(t ) is as follows: (6) In the formula, U scr is the expected charging voltage of the capacitor (V).
[0047] Rate of change of capacitor voltage error e 2( t ) is as follows: (7) For the above e 1( t ), e 2( t ), taking the derivative and organizing, the formula is obtained as follows: (8) Organizing Equation (8) into the state - space equation, we get: (9) Where: , , , . (10) (3) Construct the sliding surface of the sliding - mode prediction model according to Equations (6) - (8) s 1( t ) = 0: (11) In the formula, a 1, a 2 are the sliding coefficients of the sliding surface.
[0053] According to the sliding surface, construct the sliding - mode switching function. Considering that there are some uncertain factors in the parameters of the bidirectional DC / DC converter system, the goal is to make the tracking error e 1( t ), e 2( t ) asymptotically converge to zero. Define the sliding - mode switching function s 1( t ) as: (12) In the formula, a 1 a 2>0, is the sliding - coefficient matrix of the sliding surface;E 1( t ) is the sliding mode variable matrix.
[0054] Deriving the sliding mode switching function gives s 1 ´ ( t ): (13) (4) Reaching condition: First, it is necessary to ensure that the motion control points at random positions can converge to the sliding mode surface, but the quality of the approaching trajectory cannot be guaranteed, which may affect the stability of the system. In addition, when the system is affected by external disturbances, the quality of the sliding mode will also be affected. Therefore, it is necessary to select or design an approaching law to ensure the approaching quality of the sliding mode motion. Generally, there are several typical approaching laws for convenient calculation of the control law. They include: constant velocity approaching law, power approaching law, exponential approaching law, and general approaching law.
[0055] Define the sign function sgn s 1( t ): (14)
[0056] The form of the constant velocity approaching law is as follows: (15) Ɛ 1 represents the rate when approaching the sliding mode surface. The larger it is, the faster the reaching speed. However, too fast a speed may cause chattering of the system and is often used in scenarios such as controlling robots or drones.
[0057] The form of the power approaching law is as follows: (16) The sliding speed of the moving point towards the sliding mode surface is affected by α. In the design of the power approaching law, generally, the value is 0.5.
[0058] The form of the exponential approaching law is as follows: (17) ƞ 1 s 1( t ) is the exponential approaching term to ensure that the moving point reaches the sliding mode surface at a relatively fast rate. When the state point enters the sliding stage, the approaching speed becomes smaller at this time, and the finite-time stability of the system cannot be ensured. Therefore, an additional constant velocity approaching term - Ɛ 1 ƞ 1sgn s 1( t ) is added to the original formula. The general design principle of the parameters is to select a smaller Ɛ 1 and a larger ƞ 1.
[0059] The form of the general reaching law is as follows: (18) f ( s ) is the part to be designed, and the designer designs the corresponding reaching law according to the actual situation of the system, which also leaves some flexible elements in the design of the reaching law.
[0060] After analyzing several reaching laws, it can be seen that the design of the reaching law is to ensure the reaching rate and also pay attention to stability to avoid chattering in the system. From the control of the voltage, current steady state and transient changes of the bidirectional DC / DC converter, the exponential reaching law is more comprehensive in terms of control rate and stability. Subsequently, combined with the control theory of model state prediction, its own advantages can be enhanced, and the exponential reaching law is selected to obtain the control law of this system.
[0061] After selecting the reaching law, combining Equation (13) and Equation (18) gives: (1 - 19) After arranging Equation (19), we can get u 1 expression: (20) Adopting voltage outer loop control and current inner loop control, from we know the following relationship: (21) After integrating both sides of Equation (21) and simplifying, the expression of the controller U is obtained: (1 - 22) After designing the sliding mode controller, the stability of the system is analyzed. Select the Lyapunov function L 1 to verify the stability of the system, as shown in the following equation: (23) Taking the derivative of Equation (23) gives: (24) L 1´ is less than or equal to 0, if and only if s ( t ) = 0. This shows that the system is asymptotically stable and the system is also stable and controllable. This reaching law can design a reasonable controller, providing a theoretical basis for subsequent simulation and experiments.
[0063] (6) Construct a sliding mode surface based on the error changes of voltage and current. However, it cannot predict the future changes of the system. The model state prediction theory can predict the sliding mode switching prediction model at a future moment for the existing converter model with the expected sliding mode, and then continue to optimize the function of the sliding mode surface according to the feedback correction link and rolling optimization link of the model prediction. Finally, the controller can predict the system actions at future moments and enhance the dynamic performance of the system.
[0064] According to the sliding mode switching function s ( t ), obtain s ( t + 1) sliding mode prediction model: (25)
[0065] According to the sliding mode prediction model, the sliding mode surface, that is, the sliding mode switching function s ( t + p ) at the predicted value of the moment is as follows: (26) When the energy storage system is working, the bidirectional DC / DC converter generally shows non-linear characteristics. Some parameters may also be interfered, and its own parameters may fluctuate. There may be errors between the predicted value of the sliding mode and the actual value. Introduce a feedback correction link in the system to correct and adjust the predicted value of the sliding mode switching function at a future moment. The predicted value of t - p at the moment for t at the moment is s ( t | t - p ): (27) The predicted value after feedback correction is as shown in the following formula: (28) Add the correction coefficient h p ∈ R , as h p increases, the role of feedback correction will increase, and vice versa.
[0066] According to the predictive control theory, change the original sliding mode switching function: (29) n is the system approaching speed parameter, n > 0, sr ( t + p ) is the reference value of the sliding mode switching surface for future t + p time instants, T s is the system sampling period, nT s > 0.
[0067] According to the switching states of the switching tubes in the energy storage mode, a prediction model of the sliding mode switching function is obtained: When u 1 = 0 in Equation (21): (30) The expression of the controller changes to: (31) When u 1 = 1 in Equation (21): (32) The expression of the controller changes to: (33) After arrangement, the final control function is obtained: (34) According to the situation of the model prediction state, at t + p time instants, the inductor current values of the two states of the switching tubes corresponding to the energy storage mode are obtained, and finally converted into a duty cycle signal to control the voltage output of the DC / DC converter.
[0068] Based on the sliding mode predictive control theory, a control function is established to achieve stable control of the energy in the energy supply mode of the DC / DC converter, that is, the energy control algorithm is manifested as stable output control of the voltage in the boost output mode. The supercapacitor mode is constant power discharge, that is, the voltage of the capacitor will gradually decrease and the energy will continuously decrease, while the supercapacitor current increases to ensure constant power, that is, it matches the constant power used by the motor.
[0069] Design of the sliding mode predictive controller in the energy supply mode of the energy storage system: (1) Without considering the saturation characteristics of the inductor and the nonlinear characteristics of the output capacitor, the static voltage and current are converted into a dynamic process related to time. According to the supercapacitor U sc terminal supplies electrical energy, the common DC bus side is the load, and the load terminal resistance is R . D 2 is the duty cycle, S 2 switching function: (35) According to Kirchhoff's law, the two state equations for the system to boost voltage are as follows: (36) The space state equation of the DC / DC converter in the energy supply mode established by combining the two equations is sorted into a matrix equation: (37) (2)Adopt sliding mode predictive control for the bidirectional DC / DC converter of the supercapacitor energy storage system, combined with the expected bus voltage U dc (high-voltage output of the converter)and the inductor current i L Based on this, take the bus voltage error e 3( t ) as: (38) In the formula, U dcr is the expected bus voltage (V).
[0072] The rate of change of the bus voltage error e 4( t ) is: (39) For the above e 1( t ) and e 2( t ) are differentiated and sorted to obtain the following formula: (40) Sort the above formula (40) into a space state equation to get: (41) Among them, , , .
[0073] (3)Construct the sliding mode surface of the sliding mode prediction model according to formula (39) and formula (41) s 2( t ) = 0: (42) In the formula, a 3, a 4 are the sliding coefficients of the sliding mode surface.
[0074] According to the sliding surface, a sliding mode switching function is constructed. Considering that there are some uncertain factors in the parameters of the converter, the goal is to make the tracking error e 3( t ) e 4( t ) converge asymptotically to zero. The sliding mode switching function s 2( t ) is defined as: (43) where a 3 a 4 > 0, A 2 is the sliding coefficient matrix of the sliding surface, E 2( t ) is the sliding mode variable matrix.
[0075] Taking the derivative of the sliding mode switching function gives s 2 ´ ( t ): (44) (4) The reaching condition must first ensure that the moving point at any position can converge to the sliding surface, but it cannot guarantee the quality of the approaching trajectory, which may affect the stability of the system; the exponential reaching law is still selected to design the control law.
[0076] (45) Define the sign function sgn s 2( t ): (46) After selecting the reaching law, combining with Equation (45) gives: (47) Furthermore, the expression of u 2 can be obtained: (48) Using voltage outer loop control and current inner loop control, it can be known that: (49) After integrating both sides of Equation (48) and simplifying, the expression of the controller U is obtained: (50) (5) After designing the sliding mode controller, the stability of the system is analyzed. Select the Lyapunov function L 2 to verify the stability of the system, as shown in the following equation: (51) Deriving Equation (51) gives: (52) L 2´ is less than or equal to 0, if and only if s 2( t ) = 0. This indicates that the system is asymptotically stable and also stable and controllable. A reasonable controller can be designed based on the reaching law, providing a theoretical basis for subsequent simulations and tests.
[0077] (6) Based on the error change of voltage and current, a sliding mode surface is constructed. However, it cannot predict the future changes of the system. According to the model state prediction theory, for the existing converter model of the expected sliding mode, the sliding mode switching prediction model at a future moment can be predicted. Then, based on the feedback correction link and the rolling optimization link of the model prediction, the function of the sliding mode surface is further optimized. Finally, the controller can predict the system actions at future moments, enhancing the dynamic performance of the system.
[0078] According to the sliding mode switching function s ( t ), we obtain s ( t + 1) sliding mode prediction model: (53) According to the sliding mode prediction model, the sliding mode surface, that is, the prediction value of the sliding mode switching function s ( t + p ) at a certain moment is as follows: (54) When the energy storage system is working, the converter generally exhibits non-linear characteristics. Some parameters may be disturbed and its own parameters may fluctuate. There may be errors between the predicted values of the sliding mode and the actual values. A feedback correction link is introduced into the system to correct and adjust the predicted value of the sliding mode switching function at a future moment. The predicted value of t - p at t moment for s ( t | t - p ) is: (55) The predicted value after feedback correction is as follows: (56) Adding the correction coefficient h p ∈ R , as h pAs the
[0079] increases, the role of feedback correction increases, and vice versa. (57) n is the system approaching speed parameter, n > 0, s r ( t + p ) is the reference value of the sliding mode switching surface at the future t + p moment, T s is the system sampling period, nT s > 0.
[0080] According to the switching state of the switch tube under the power supply mode, a prediction model of the sliding mode switching function is obtained: When u 2 = 0 in Equation (49): (58) The expression of the controller changes to: (59) When u 2 = 1 in Equation (49): (60) The expression of the controller changes to: (61) After sorting, the final control function is obtained: (62) According to the model prediction state of the above formula, at t + p moment, the inductor current values of the two states of the switch tube under the corresponding power supply mode are obtained, and finally converted into a duty ratio signal to control the voltage output of the bidirectional DCDC converter.
[0081] The mathematical model of the energy self - coordinated control strategy of the energy storage system is established, that is, the core controller part is built. According to parameters such as the bus voltage, supercapacitor voltage, and hydraulic pad displacement, the energy storage system selects a mode. According to the selected mode, plus parameters such as the desired bus voltage, desired supercapacitor terminal voltage, and current, the controller is calculated. The final calculation result is converted into a duty ratio signal that the converter can execute to complete the effective control of the output current and voltage, that is, to complete the stable control of energy; thus, the energy self - coordinated control strategy of the pump - controlled hydraulic pad energy storage system is established.
Claims
1. A method for self-coordinating energy control of a pump-controlled hydraulic cushion energy storage system, characterized in that: The energy storage system has three modes: energy storage mode, standby mode and energy supply mode. The energy self-coordination control method of the energy storage system includes the following processes: S1. Design an energy self-coordination control strategy for a pump-controlled hydraulic pad energy storage system, wherein the energy self-coordination control strategy includes multi-condition switching of energy storage or energy supply logic and stable input or output control of energy; S2, multi-condition switching of energy storage or energy supply logic: switch the working mode of the energy storage system according to the change of DC bus voltage of the energy storage system, the displacement of the mold pad of the pump-controlled hydraulic pad, and the change of the supercapacitor of the energy storage system; S3. Stable output or input control of energy: The energy storage mode is used to manage the inductor current and capacitor voltage on the energy storage side of the bidirectional DC / DC converter, and the energy supply mode is used to regulate the inductor current and bus voltage on the bus side of the bidirectional DC / DC converter. Both the energy storage mode and the energy supply mode are precisely coordinated through the sliding mode predictive control algorithm.
2. A method for self-coordinating energy control of a pump-controlled hydraulic cushion energy storage system as claimed in claim 1, characterized in that: The sliding mode predictive control algorithm includes: S31, establishing a state space equation of the bidirectional DC / DC converter in energy storage mode and energy supply mode; S32, constructing a sliding surface according to the desired output / input voltage and inductor current; S33, determining the sliding mode switching function according to the sliding mode surface, and combining the feedback correction and rolling optimization ideas of the model predictive control to achieve optimal control of the switching function; S34, select a suitable sliding mode approach law to ensure that the system reaches the sliding mode surface quickly and stably; S35, select and substitute into calculation according to the algorithm formula, and finally solve the expression of the controller; S36. Verify the stability of the system through Lyapunov function.
3. A method for self-coordinating energy control of a pump-controlled hydraulic cushion energy storage system as claimed in claim 2, characterized in that: The energy storage mode, standby mode and energy supply mode cover four working stages of the pump-controlled hydraulic pad, which include the pre-acceleration stage in which the die pad moves downward, the stretching stage in which the die pad is pushed downward, the stage in which the blank holding force is maintained, and the stage in which the die pad moves upward and returns.
4. A method for self-coordinating energy control of a pump-controlled hydraulic cushion energy storage system as claimed in claim 3, characterized in that: During the stage of maintaining the blank holding force, the upward return stage of the die pad, the pre-acceleration stage, and the stretching stage, the energy storage system monitors the DC bus voltage and selects to enter the standby mode, energy storage mode, or energy supply mode according to the DC bus voltage. If the voltage is less than the specified range, the energy supply mode is started; if it is greater than the specified range, the energy storage mode is started; if the voltage is within the specified range, the standby mode is entered; When the charge value is not between 0 and the maximum value, the safety system is entered, the braking resistor protection in the safety system is started, and the regenerative energy in the working cycle is consumed by the braking resistor.
5. A method for self-coordinating energy control of a pump-controlled hydraulic cushion energy storage system as claimed in claim 4, characterized in that: The implementation steps of the sliding mode predictive controller of the energy storage system in the energy storage mode include: Convert static voltage and current into dynamic equations related to time, and define S The switch function of 1 is as follows: ； in: T is the switching cycle, D 1 is the duty cycle; According to Kirchhoff's law, the state equation of the energy storage system in the energy storage mode is: ； In the formula, is the voltage of the supercapacitor bank; is the load terminal voltage; i L is the inductor current; is the equivalent resistance of the supercapacitor; is the inductor; The established spatial state equation of the bidirectional DC / DC converter in the energy storage mode is organized into a matrix equation: ； The bidirectional DC / DC converter of the supercapacitor energy storage system adopts sliding mode predictive control, combined with the capacitor charging voltage to be controlled U sc and inductor current i L Based on the capacitor charging voltage error e 1( t )for: ； In the formula, U scr is the expected charging voltage of the capacitor; Capacitor voltage error change rate e 2( t )for: ； To the top e 1( t ), e 2( t ) is derived and sorted to get the following formula: ； Arranged into the spatial state equation, we get: ； in: ， ， ， ； Constructing the sliding surface of the sliding mode prediction model s 1( t )=0: ； In the formula, a 1. a 2 is the sliding coefficient of the sliding surface; According to the sliding surface, the sliding mode switching function is constructed. s 1( t ) is defined as: ； In the formula, a 1 a 2>0, is the sliding coefficient matrix of the sliding surface; E 1( t ) is the sliding mode variable matrix; Derivative processing of the sliding mode switching function yields s 1 ´ ( t ): ； Define the symbolic function sgn s 1( t ): ； The form of the isokinetic reaching law is as follows: ； 1 represents the velocity approaching the sliding surface; The form of the power reaching law is as follows: ； The form of the exponential reaching law is as follows: ； The general reaching law has the following form: ； From this we get: ； Then get u 1 Expression: ； Adopt voltage outer loop control and current inner loop control. Know: ； Get the controller U The expression is: ； Select Lyapunov function L 1Verify the stability of the system as follows: ； The derivative is: ； According to the sliding mode switching function s ( t ),get s ( t +1) Sliding mode prediction model: ； According to the sliding mode prediction model, the sliding surface is also the sliding mode switching function s ( t + p ) is as follows: ； Introducing a feedback correction link into the system, the sliding mode switching function prediction value at a certain moment in the future is corrected and adjusted. t - p Always t The predicted value at time is s ( t | t - p ): ； The predicted value after feedback correction is as follows: ； Add correction factor h p ∈ R ,along with h p As increases, the effect of feedback correction will increase, and vice versa; According to the predictive control theory, the original sliding mode switching function is changed: ； n is the system approach speed parameter, n >0, s r ( t + p ) for the future t + p The reference value of the sliding mode switching surface at the moment, T s is the system sampling period, nT s >0; when u When 1=0: ； The controller expression changes to: ； when u When 1=1: ； The controller expression changes to: ； The final control function is obtained by sorting: ； A control function is established based on the sliding mode predictive control theory to achieve stable energy control of the bidirectional DC / DC converter in energy storage mode.
6. A method for self-coordinating energy control of a pump-controlled hydraulic cushion energy storage system as claimed in claim 5, characterized in that: The implementation steps of the sliding mode predictive controller of the energy storage system in the energy supply mode include: Converting static voltage and current into a time-dependent dynamic process, supercapacitors U sc The end provides power, and the common DC bus side is the load. S The switch function of 2 is as follows: ； in: R is the load terminal resistance, D 2 is the duty cycle; According to Kirchhoff's law, the state equation of the energy storage system in the energy supply mode is: ； The established spatial state equation of the bidirectional DC / DC converter in the energy supply mode is organized into a matrix equation: ； The bidirectional DC / DC converter of the supercapacitor energy storage system adopts sliding mode predictive control, combined with the bus expected voltage U dc and inductor current i L Based on the bus voltage error e 3( t )for: ； In the formula, U dcr is the expected bus voltage; Bus voltage error change rate e 4( t )for: ； To the top e 1( t ), e 2( t ) is derived and sorted to get the following formula: ； Arranged into the spatial state equation, we get: ； in: ， ， ； Constructing the sliding surface of the sliding mode prediction model s 2( t )=0: ； In the formula, a 3. a 4 is the sliding coefficient of the sliding surface; According to the sliding surface, the sliding mode switching function is constructed. s 2( t ) is defined as: ； In the formula, a 3 a 4>0, A 2 is the sliding coefficient matrix of the sliding surface, E 2( t ) is the sliding mode variable matrix; Derivative processing of the sliding mode switching function yields s 2 ´ ( t ): ； ； Define the symbolic function sgn s 2( t ): ； After the reaching law is selected, we get: ； Then get u 2 Expression: ； Using voltage outer loop control and current inner loop control, we can know: ； After integrating both sides and simplifying, we get the controller U The expression is: ； The Lyapunov function is selected to verify the stability of the system, as shown below: ； The derivative is: ； According to the sliding mode switching function s ( t ),get s ( t +1) Sliding mode prediction model: ； According to the sliding mode prediction model, the sliding surface is also the sliding mode switching function s ( t + p ) is as follows: ； Introducing a feedback correction link in the energy storage system, the predicted value of the sliding mode switching function at a certain moment in the future is corrected and adjusted. t - p Always t The predicted value at time is s ( t | t - p ): ； The predicted value after feedback correction is as follows: ； Add correction factor h p ∈ R ,along with h p As increases, the effect of feedback correction will increase, and vice versa; According to the predictive control theory, the original sliding mode switching function is changed: ； n is the system approach speed parameter, n >0, s r ( t + p ) for the future t + p The reference value of the sliding mode switching surface at the moment, T s is the system sampling period, nT s >0; The prediction model of the sliding mode switching function is obtained: when u When 2=0: ； The controller expression changes to: ； when u When 2=1: ； The controller expression changes to: ； The final control function is obtained by sorting: ； A control function is established based on the sliding mode predictive control theory to achieve stable energy control of the bidirectional DC / DC converter in the energy supply mode.