Energy-feeding suspension control system based on PI feedforward model predictive control algorithm

Abstract

The present invention discloses a feed-back type suspension control system based on a PI feed-forward model predictive control algorithm, which belongs to the field of vehicle control and includes the following steps: S1, establishing a dynamic model of a feed-back suspension system; S2, establishing a simulation model of an energy feedback control system; S3, designing a voltage outer loop controller, calculating a reference current through a PI controller and a model-based feed-forward compensator; and designing a load disturbance observer to observe changes in the load; S4, designing a current inner loop controller, calculating an optimal duty cycle through a model predictive control algorithm; and designing an inductor current observer to perform an equivalent estimation of the inductor current to avoid the influence of changes in input voltage and inductor resistance on the quality of the feedback power. The present invention can effectively improve the quality of the feedback power, and at the same time, can still maintain a good control effect when external disturbances occur. In addition, the present invention can also effectively improve the ride comfort and handling stability of the vehicle.

Description

Technical Field

[0001] The invention relates to the field of vehicle control, and in particular to an energy-feedback suspension control system based on a PI feedforward model predictive control algorithm. Background Art

[0002] As environmental and energy issues become increasingly severe, the research and development and improvement of automobile energy-saving technologies have become a hot topic worldwide. The suspension is an important component in the car, mainly used to absorb and attenuate the vibration caused by uneven road surfaces. Converting the vibration energy of the suspension into electrical energy and recycling it has a very significant effect on improving the fuel economy of the car. Therefore, how to efficiently and reliably convert the vibration energy of the suspension into usable electrical energy has become the research focus of scholars at home and abroad.

[0003] In recent years, researchers have conducted extensive research on the actuator configuration, energy recovery efficiency and dynamic performance of energy-feeding suspension. Energy-feeding suspension can be divided into rotary motor type and linear motor type according to the different actuator configurations. The rotary motor type requires a transmission mechanism to convert the up and down vibration of the suspension into rotational motion, but this type of motor has a complex structure and low working efficiency, while the linear motor can directly convert the vibration of the suspension, with the advantages of simple installation and less mechanical loss. Therefore, linear motors are widely used in the research of energy-feeding suspension. In order to solve the problem of insufficient vibration energy recovery, researchers have adopted a new auxiliary power balancing strategy and designed a control system to overcome damping failure to improve the energy conversion efficiency of the suspension and improve the dynamic performance of the suspension.

[0004] With the rapid development of Internet of Vehicles technology, fields such as intelligent transportation and autonomous driving require the use of multiple sensors to monitor road and vehicle conditions, and the requirements for power quality are becoming more and more stringent. It is of great significance to design an effective method to improve the quality of feedback power and realize direct power supply to on-board electrical equipment such as sensors. In the energy feedback control system, the Boost circuit plays a very important role in the process of energy conversion. Therefore, the present invention mainly focuses on the control strategy of the Boost circuit. At the same time, considering that the disturbance of the parameters in the energy feedback control system may affect the effect of energy feedback, the present invention designs a disturbance observer to observe the uncertainties in the system, thereby improving the robustness of the energy feedback control system. The method proposed in the present invention can convert the random and irregular alternating current generated by vibration into stable and reliable direct current, so as to achieve the purpose of directly powering on-board electrical equipment. Summary of the invention

[0005] In order to solve the technical problems mentioned in the above background technology, the present invention proposes a feed-back suspension control system based on a PI feedforward model predictive control algorithm, which can convert the electrical energy generated by suspension vibration into stable and reliable direct current for use by vehicle-mounted electrical equipment.

[0006] In order to achieve the above technical objectives, the technical solution of the present invention is:

[0007] The energy-feedback suspension control system based on the PI feedforward model predictive control algorithm includes:

[0008] A suspension dynamics system, comprising a spring, a damping shock absorber and an electromagnetic linear motor actuator, wherein the suspension dynamics system is used to convert the vibration energy of the vehicle during driving into irregular alternating current;

[0009] The energy feedback control system comprises a rectifier filter module, a boost module and a PI feedforward model predictive control module, wherein the energy feedback control system is used to convert irregular alternating current into stable and reliable direct current, wherein the PI feedforward model predictive control module comprises a PI feedforward control module and a model predictive control module.

[0010] Optionally, the electromagnetic linear motor actuator is used to generate relative movement between the permanent magnet and the coil in the motor when the vehicle is traveling on an uneven road, thereby generating an induced electromotive force and an induced current.

[0011] Optionally, the electromagnetic linear motor is used to generate electromagnetic damping force to hinder the relative movement of the suspension according to the electromagnetic damping principle, thereby attenuating the vibration caused by uneven road surface and improving the ride comfort of the vehicle.

[0012] Optionally, the rectification and filtering module is used to convert irregular alternating current into direct current.

[0013] Optionally, the energy feedback control system further includes: the boost module is used to increase the amplitude of the energy feedback voltage, thereby avoiding the voltage amplitude being too small to be used.

[0014] Optionally, the energy feedback control system also includes: the PI feedforward control module includes a current inner loop and a voltage outer loop, the voltage outer loop uses a PI controller combined with feedforward compensation to calculate a reference current, and the current inner loop outputs an optimal duty cycle through a model predictive control module, driving the pulse width modulator to generate a pulse signal to control the on and off of the MOSFET, so that the energy feedback voltage is stabilized at the desired voltage value.

[0015] Optionally, the model of the PI feedforward control module is:

[0016]

[0017] Where: i ref is the reference current, u ref is the reference voltage, is the observed value of the disturbance variable, is the observed value of the load equivalent resistance, is the observed value of output voltage.

[0018] Optionally, the model of the model predictive control module is:

[0019]

[0020] Where: d is the duty cycle, The observed value of the disturbance variable, is the output voltage observation value, i ref is the reference current, is the observed value of the inductor current, L is the inductor value, and T is the sampling period.

[0021] Optionally, the energy feedback control system further includes:

[0022] The inductor current disturbance observer is used to estimate the inductor current in the boost module as the input of the current inner loop to avoid the influence of the disturbance of the two on the energy feedback effect, wherein the boost module is the Boot circuit; the load disturbance observer is used to observe the change of the load to avoid affecting the stability of the energy feedback voltage when the load changes suddenly. The energy feedback control system can respond quickly to the disturbance through the observer's estimation of the load resistance, thereby improving the robustness and dynamic response speed of the energy feedback control system.

[0023] Due to the adoption of the above technical solution, the technical progress achieved by the present invention is:

[0024] The energy feedback suspension control system disclosed in the present invention includes: a suspension dynamics system and an energy feedback control system. Optionally, the suspension dynamics system includes: a spring, a damping shock absorber, and an electromagnetic linear motor actuator. The energy feedback control system includes: a rectifier filter module, a boost module and a control module. The control module in the energy feedback control system adopts a PI feedforward model predictive double closed-loop control algorithm. The voltage outer loop uses PI feedforward control to limit and control the current to obtain a reference current, and the current inner loop uses model predictive control to output the optimal duty cycle, so that the energy feedback voltage can quickly and accurately track the expected value. The system takes into account the impact of unknown parameters and disturbances on the system, and designs a disturbance observer to observe uncertainties in the system, thereby improving the robustness and dynamic characteristics of the system. The suspension dynamics system improves the dynamic performance of the system because the electromagnetic linear motor actuator will generate electromagnetic damping force. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0026] Figure 1 This is a schematic diagram of the energy feedback suspension according to an embodiment of the present invention;

[0027] Figure 2 This is a diagram of the dynamic model of a quarter energy-feeding suspension according to an embodiment of the present invention;

[0028] Figure 3 This is a topology diagram of a Boost circuit according to an embodiment of the present invention;

[0029] Figure 4 It is an overall block diagram of the PI feedforward model predictive control model according to an embodiment of the present invention;

[0030] Figure 5 This is a waveform diagram of the inductor current in one cycle of an embodiment of the present invention;

[0031] Figure 6 This is a simulation diagram of random road surface excitation according to an embodiment of the present invention;

[0032] Figure 7 This is a diagram of the energy feeding voltage and ripple waveform under the PI control algorithm of an embodiment of the present invention;

[0033] Figure 8 This is a diagram of the energy feeding voltage and ripple waveform under the PI feedforward model predictive control algorithm of an embodiment of the present invention;

[0034] Fig. 9 This is a waveform diagram of the transient response of the energy feeding voltage when the load changes without feedforward control in the embodiment of the present invention;

[0035] Fig.10 This is a waveform diagram of the transient response of the energy feeding voltage when the load changes under the PI feedforward model predictive control according to the embodiment of the present invention;

[0036] Fig.11 The sprung acceleration simulation diagrams of different systems in the embodiments of the present invention;

[0037] Fig.12 The tire dynamic load simulation diagram under different systems of the embodiment of the present invention;

[0038] Fig.13 It is a simulation diagram of the suspension dynamic travel under different systems of the embodiments of the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] The purpose of the present invention is to provide a feed-back suspension control system based on a PI feed-forward model predictive control algorithm, which can effectively improve the quality of the feedback power, thereby solving the problem that the power quality generated by the suspension vibration is poor and cannot directly supply power to the on-board electrical equipment. At the same time, it also improves the dynamic performance of the suspension to a certain extent.

[0041] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0042] Figure 1 FIG. 1 is a schematic diagram of the energy-feeding suspension according to an embodiment of the present invention. Figure 1 As shown, the energy feedback suspension consists of a suspension dynamics system and an energy feedback control system.

[0043] The suspension dynamics system is mainly composed of springs, damping shock absorbers, electromagnetic linear motor actuators, etc.

[0044] The energy feedback control system mainly includes a rectifier and filter module, a boost module and a control module.

[0045] The suspension dynamics system can convert the vibration energy of the vehicle during driving into irregular alternating current.

[0046] The energy feedback control system can convert irregular alternating current into stable and reliable direct current.

[0047] The rectification and filtering module can convert irregular alternating current into direct current.

[0048] The boost module can increase the amplitude of the rectified DC voltage, thereby preventing the voltage from being too small to be used.

[0049] The control module can stabilize the energy feeding voltage at a desired value, thereby improving the quality of the feedback electric energy.

[0050] Figure 2 1 / 4 energy-feeding suspension dynamics model of the present invention. Figure 2As shown, the vehicle model is converted into a dynamic model. The present invention makes the following assumptions: (1) the vehicle is symmetrical; (2) the influence of the front and rear vibrations of the vehicle is ignored; (3) the damping of the tires is ignored and is equivalent to a spring; (4) the mass of the energy feedback device is small compared to the mass of the vehicle body and wheels and can be ignored.

[0051] The 1 / 4 energy-feeding suspension dynamics model can obtain the dynamics equation as follows according to Newton's law:

[0052]

[0053] Where: m s is the sprung mass; m u is the unsprung mass; z s is the sprung mass displacement; z u is the unsprung mass displacement; z r is the vertical displacement of the road surface; k s is the suspension stiffness; k t is the tire stiffness; c s is the suspension damping coefficient; F is the force.

[0054] Select state variables have to

[0055]

[0056] in,

[0057]

[0058] The dynamic performance evaluation indicators of the 1 / 4 energy feedback suspension system should include:

[0059] (1) Ride comfort

[0060] Ride comfort is the acceleration of the vehicle body It refers to the feeling of suspension vibration when the vehicle is driving on uneven roads. The smaller the vehicle acceleration, the higher the ride comfort.

[0061] (2) Handling stability

[0062] Handling stability is the dynamic load on the tire z u -z r It means that in order to ensure driving safety, the tires should maintain contact with the road surface during driving. The smaller the dynamic load of the tires, the higher the vehicle's operating stability.

[0063] (3) Vehicle posture

[0064] Body posture, i.e. suspension travel z s -z uIt refers to the relative displacement of the sprung mass and the unsprung mass. If the suspension travel is too large during vehicle driving, the axle will hit the limit block, which will reduce the service life of the vehicle.

[0065] (4) Energy Feedback Potential

[0066] The energy feedback potential is related to the relative speed between the suspension sprung mass and the unsprung mass. The greater the suspension vibration speed, the greater the induced electromotive force generated by the generator and the more electrical energy generated, and vice versa.

[0067] The present invention comprehensively evaluates the dynamic performance of the energy feedback suspension through the above four indicators.

[0068] When the vehicle is driving on an uneven road, the vibration of the suspension will cause relative movement between the permanent magnet and the coil in the electromagnetic linear motor. According to Faraday's law of electromagnetic induction, the electromagnetic linear motor will generate an induced electromotive force and an induced current as follows:

[0069]

[0070] Where: u e is the induced electromotive force; i e is the induced current; k e is the back electromotive force coefficient of the linear motor; R is the equivalent resistance of the energy feeding circuit; r is the internal resistance of the linear motor.

[0071] At the same time, according to the electromagnetic damping principle, the electromagnetic linear motor will generate a damping force to hinder the relative movement of the suspension. The electromagnetic damping force is:

[0072] F = k i i e (5)

[0073] Where: k i is the thrust coefficient of the linear motor.

[0074] Combining equations (4) and (5), we can obtain:

[0075]

[0076] make At this time, the generator can be equivalent to a damping shock absorber, and the electromagnetic damping force can be written as:

[0077]

[0078] Therefore, the energy-feeding suspension will generate an equivalent electromagnetic damping force to attenuate the vibration caused by the uneven road surface while recovering energy, thereby improving the dynamic performance of the vehicle suspension system.

[0079] Figure 3is the Boost circuit topology of the embodiment of the present invention. Wherein: i ,u o are the input and output voltages of the Boost circuit, R is the load equivalent resistance, and R L is the inductor resistance, L is the inductor inductance, D is the diode, S is the controllable switch tube, and C is the output capacitor. In most cases, the vibration amplitude of the suspension is not large, which will cause the voltage generated by the generator to be small, thus affecting the effect of energy feedback. The Boost circuit can perform a voltage-boosting conversion on the rectified electric energy, and stabilize the amplitude of the energy-feeding voltage at the desired value by controlling the on-off time of the switch tube.

[0080] Since the common voltage of vehicle-mounted electrical equipment is 24V, this article sets the expected voltage value to 24V.

[0081] The Boost circuit has two switching states:

[0082] When the switch tube S of the boost circuit is turned on, the diode D is in the cut-off state, and the input voltage u i The inductor L is charged. The inductor L stores energy due to the conduction of the switch tube S. The load R is powered by the capacitor C. The inductor current i can be obtained. L With output voltage u o The equation is:

[0083]

[0084] When the switch tube S of the boost circuit is turned off, the diode D is in the on state, and the inductor L releases energy due to the turn-off of the switch tube S. The input voltage u i The inductor L supplies power to the load R and the capacitor C at the same time, and the inductor current i L With the voltage equation u o for:

[0085]

[0086] Select the inductor current i L (t) and output voltage u o (t) is the state variable, d(t) is the control input. According to the state space averaging method, the state space expression of the Boost circuit in the continuous current mode can be obtained as follows:

[0087]

[0088] in:

[0089]

[0090] Figure 4 FIG. 1 is an overall block diagram of the PI feedforward model predictive control model according to an embodiment of the present invention. Figure 4 As shown, the PI feedforward model predictive control is divided into voltage outer loop control and current inner loop control.

[0091] The model predictive control model in the PI feedforward model predictive control system can speed up the dynamic response speed of the control system, but at the same time it will bring steady-state errors. The integral link of the PI feedforward control can eliminate the steady-state errors in the system, and the two complement each other. Therefore, the present invention combines the PI feedforward control algorithm with the model predictive control algorithm.

[0092] Some parameters in the PI feedforward model predictive control system model are unknown or have disturbances. These inaccurate parameters will greatly affect the quality of the feedback power. To this end, the present invention designs a disturbance observer to estimate these parameters. The specific system parameters involved include the input voltage u i 、Inductor resistance R L and load equivalent resistance R, etc. In the voltage outer loop, the equivalent load resistance R will change due to the different number of electrical devices in use. Therefore, the load disturbance observer is designed in the present invention to observe the change of the load to avoid the influence of the load disturbance on the feeding effect, and then the PI feedforward controller is used to generate a response to the voltage deviation, so as to obtain the reference current; in the current inner loop, considering that the changes of the input voltage and the inductor resistance in the Boost circuit may affect the inductor current, the inductor current observer is designed in the present invention to estimate the inductor current, and then the duty cycle is calculated by the model predictive control algorithm according to the relationship between the current output by the inductor current observer and the reference current, and then the pulse width modulator is driven to generate a pulse signal to control the on and off of the MOSFET.

[0093] Figure 5 FIG. 1 is a waveform diagram of the inductor current in one cycle of an embodiment of the present invention. Figure 5 As shown in the figure, the predictive current mode control is that when each control cycle starts, the input voltage u i , output voltage u o and the inductor current i L Sampling is performed, and then the rising rate and falling rate of the inductor current are calculated based on the sampled values. Since the switching cycle of the Boost circuit is very short, the slope of the inductor current in two adjacent switching cycles can be approximately considered to remain unchanged. Therefore, the current at the next moment can be predicted by the slope of the inductor current waveform.

[0094] The slope calculation formula of the inductor current in each stage is as follows:

[0095]

[0096] Depend on Figure 5 It can be seen that the sampling current i at time kT LThe duty cycle d(k+1) and the rising and falling rates of the inductor current in the (k) and (k+1)T cycles can be used to calculate the inductor current i in the next cycle. L (k+1) is:

[0097] i L (k+1)=i L (k)+k1t1+k2t2+k3t3 (12)

[0098] Since t2=d(k+1)T, combining equation (11) and equation (12), we can get:

[0099]

[0100] Let i ref =i L (k+1), the duty cycle d(k+1) under the predicted current mode control can be obtained from formula (13):

[0101]

[0102] Due to the influence of factors such as road roughness and changes in motion state, the vibration of the suspension will be very accidental, random and uncertain, which will cause the voltage generated by the generator to change continuously. At the same time, there is also disturbance in the inductor internal resistance in the energy feedback control system. The inductor current observer is used to monitor the input voltage u i And the inductor resistance R L By making an estimate, the sudden change of the inductor current caused by the two disturbances can be avoided, thereby improving the quality of the feedback electric energy.

[0103] When the input voltage u i And the inductor internal resistance R L When the change occurs, the error between the two has an influence on the inductor current i L The impact is shown in formula (15):

[0104]

[0105] Define the input voltage u i and the inductor internal resistance R L The disturbance variable caused is:

[0106] w(t)=u i (t)-i L (t)R L (16)

[0107] Since the control system is generally operated at a high frequency, w(t) is approximately unchanged within a cycle, so we can get:

[0108] w(k+1)=w(k) (17)

[0109] Substituting the disturbance variable into equation (16) and using the Euler algorithm to discretize the inductor current, we can obtain:

[0110]

[0111] The inductor current i L (k) and disturbance variable w(k) are state variables, and equations (17) and (18) are written in matrix form as follows:

[0112]

[0113] Where:

[0114]

[0115] Therefore, the inductor current observer can be designed as:

[0116]

[0117] in: is the observed value of the inductor current, is the input voltage u i and the inductor resistor R L The disturbance observed.

[0118] Combining (20) and (21), we can get the error between the state variable and the observed value:

[0119]

[0120] To make the system stable, it is required that (A L -HC L ) must be a Hurwitz matrix, and the observed variable formula is:

[0121]

[0122] The present invention obtains the disturbance variable by using the inductor current observer It is not necessary to use the input voltage u directly i and the inductor resistor R L value, we can get the inductor current i at the next moment L (k+1), which can effectively avoid the impact of the disturbances of the two on the feedback power quality.

[0123] The current inner loop control adopts the model predictive control algorithm. First, based on the established discrete model and the final desired goal, the minimum value of the objective function at that moment is obtained under the premise of satisfying the restriction conditions to obtain the optimal switching state. The latest data fed back in each cycle is continuously refreshed, and the optimization problem after refreshing is solved, so as to finally realize the iterative optimization of the model.

[0124] The whole process of the model predictive control algorithm adopted by the current inner loop mainly includes the following three steps:

[0125] (1) Establish an equivalent model for the inductor current prediction value. Combining equations (18) and (23), the predicted value of the inductor current at time k+1 after using the observer is obtained as follows:

[0126]

[0127] (2) Establishing the target cost function. The output voltage u is considered in the present invention. o and the inductor current i L There are two error terms. Therefore, in order to facilitate the solution of the expression, it is more conducive to finding the maximum value of the objective function to use the quadratic error. Therefore, the objective function is set as:

[0128]

[0129] (3) The error between the reference current and the predicted current at the next moment is evaluated and analyzed through the objective function, and the duty cycle that minimizes the objective function is selected and acts on the switch tube:

[0130]

[0131] From equation (26), it can be seen that in order to calculate the optimal duty cycle d(k+1) of the energy feedback control system, it is also necessary to obtain the observed value of the output voltage through the voltage loop: And the reference current i ref .

[0132] The voltage outer loop control uses a PI controller and a model-based feedforward compensator to generate a response to the voltage deviation, thereby obtaining a reference current. In order to avoid the impact of changes in the load equivalent resistance R on the energy feedback control system, a load disturbance observer is used to observe the disturbance of the equivalent resistance R, and a feedforward compensation controller is used to quickly respond to the disturbance, which can effectively improve the robustness and dynamic response characteristics of the energy feedback control system. Therefore, the reference current is mainly composed of two parts, one part is calculated by the observer estimate, and the other part is obtained by correcting the steady-state error through the PI controller in the voltage outer loop. Therefore, the reference current can be expressed as:

[0133] i ref (k+1)=iref·ob (k+1)+i ref·co (k+1) (27)

[0134] According to the system power balance principle, we can get:

[0135]

[0136] Combining equations (27) and (28), the observer output reference current can be obtained as:

[0137]

[0138] When different electrical devices are connected, the load equivalent resistance R will change, and this change is unknown. In order to improve the robustness of the energy feedback control system against load equivalent resistance disturbances, a load disturbance observer is designed to estimate the load equivalent resistance.

[0139] According to the Boost circuit model, the current relationship can be obtained as shown in formula (30):

[0140]

[0141] Where: i D ,i C ,i o are the currents flowing through the diode, capacitor and load resistor respectively.

[0142] Rearranging formula (30) yields:

[0143]

[0144] With output current i o (t), output voltage u o (t) Establish the state space equation for the state variable and discretize it using the Euler algorithm to obtain:

[0145]

[0146] in:

[0147]

[0148] Therefore, the load disturbance observer can be designed as:

[0149]

[0150] in:

[0151]

[0152] are the observed values ​​of the current flowing through the load and the voltage across the load, respectively.

[0153] Combining equation (32) and equation (33), we can get:

[0154]

[0155] To make the system stable, it is required that (A o -HC o ) must be a Hurwitz matrix, then the observed values ​​of the output variables and the observed values ​​of the load can be obtained as shown in formula (35):

[0156]

[0157] In addition, another part of the reference current is calculated by the PI controller in the voltage outer loop:

[0158]

[0159] The reference current required by the inner current loop can be obtained through equation (36). The change of equivalent resistance can be observed through the load disturbance observer, and feedforward compensation control can be used to respond quickly to disturbances, thereby improving the robustness and dynamic response characteristics of the system and effectively improving the quality of the feedback power.

[0160] Figure 6 This is a simulation diagram of random road excitation in an embodiment of the present invention. In order to verify the universal applicability of the method proposed in the present invention, a road surface roughness signal is generated by using a filtered white noise method to simulate the actual situation of a random road surface, and the road surface speed expression can be obtained as follows:

[0161]

[0162] Where: f0 is the lower cutoff frequency; n0 is the reference spatial frequency; G q (n0) is the road roughness coefficient; w(t) is the time domain signal of unit white noise.

[0163] In the present invention, random road surface excitation with a road surface grade of C and a vehicle speed of 20 m / s is selected for simulation to prepare for verifying the effectiveness of the method proposed in the present invention.

[0164] Figure 7 This is a diagram of the energy feeding voltage and ripple waveform under the PI control algorithm of an embodiment of the present invention.

[0165] The PI control algorithm is based on k p , k i The adjustment of the two parameters completes the design of the control rate. The new control amount is calculated based on the difference between the expected value and the feeding voltage, so that the feeding voltage can follow the expected value.

[0166] Figure 8 This is a waveform diagram of the energy feeding voltage and ripple under the PI feedforward model predictive control according to an embodiment of the present invention.

[0167] The PI feedforward model predictive control adds a current inner loop to the voltage single loop to form a double closed-loop control. The inductor current is controlled by setting an observer and a feedforward compensator to further improve the performance of the energy feedback control system. The outer loop uses a method combining PI control and feedforward compensation to calculate the reference current, and the inner loop uses the model predictive control algorithm to output the optimal duty cycle when it approaches the desired voltage. The objective function of the model predictive control is expressed by a quadratic error expression. Since single-step MPC and multi-step MPC perform equally well in practical applications, and considering the computational burden of the processor, this paper adopts a single-step prediction method.

[0168] After completing the model building of the energy feedback suspension dynamics system and the energy feedback control system, simulation experiments were carried out on the two control methods under the same simulation conditions.

[0169] Depend on Figure 7 and Figure 8 It can be seen that both control algorithms can convert the recovered electric energy into stable direct current. However, compared with the voltage single-loop PI control algorithm, the PI feedforward model predictive control has a better performance, effectively reducing the ripple of the feeding voltage, and has a faster regulation speed and smaller overshoot, thereby improving the quality of the feedback electric energy and achieving the purpose of directly powering the on-board electrical equipment.

[0170] Therefore, the energy feedback control system is controlled by the PI feedforward model predictive control algorithm, which successfully converts the random and irregular alternating current generated by the suspension vibration into continuous and stable direct current, improving the quality of the feedback electric energy, which is of great significance to the energy recovery and reuse of the energy-feeding suspension.

[0171] Fig. 9 and Fig.10 The figure is a transient response waveform image of the energy feedback voltage when the load equivalent resistance changes under the two algorithms of feedforward control and PI feedforward model predictive control in the embodiment of the present invention. In order to verify that the energy feedback control system proposed in the present invention has good adaptability to the equivalent load resistance disturbance, the load resistance is changed from 20Ω to 10Ω at 0.03s.

[0172] Depend on Fig. 9 and Fig.10It can be seen that when the load equivalent resistance changes, both control algorithms can make the energy feeding voltage return to a stable state again. However, the dynamic response speed of the model predictive control algorithm with feedforward control is faster. Without feedforward control, the energy feeding voltage has an amplitude drop of about 2.1%, while the amplitude drop is less than 0.38% under the PI feedforward model predictive control algorithm. This shows that the algorithm proposed in the present invention can better cope with the disturbance of the equivalent load resistance and effectively improve the robustness of the energy feedback control system.

[0173] Figure 11-13 The sprung acceleration, tire dynamic load, and suspension dynamic stroke simulation diagrams under different systems of the present invention are shown respectively. The electromagnetic linear motor actuator can generate electromagnetic damping force to adjust the dynamic performance of the suspension while recovering energy.

[0174] In order to verify the improvement effect of the method proposed in the present invention on the suspension dynamics performance, the energy feedback suspension system in the present invention and the passive suspension system with the same parameters are simulated under random road excitation.

[0175] like Figure 11-13 As shown, the red line represents the simulation curve of the energy feedback suspension, and the blue line represents the simulation curve of the passive suspension.

[0176] Compared with the passive suspension with the same parameters, the energy-feeding suspension of the present invention has improvements in sprung acceleration and tire dynamic load, indicating that the energy-feeding suspension has better ride comfort and handling stability. However, the suspension dynamic travel has deteriorated, so when designing the energy-feeding suspension, consideration should be given to increasing the suspension's movement space to prevent the vehicle from hitting the limit block during driving, which would deteriorate the vehicle's ride comfort.

[0177] The present invention uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only used to help understand the method and core ideas of the present invention. At the same time, for those skilled in the art, according to the ideas of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. The energy-feedback suspension control system based on the PI feedforward model predictive control algorithm is characterized by: include: The 1 / 4 energy-feeding suspension dynamics model, linear motor and Boost circuit model are established. The linear motor is used to generate relative movement between the permanent magnet and the coil in the motor due to the vibration of the suspension when the vehicle is driving on an uneven road, thereby generating induced electromotive force and induced current, and generating electromagnetic damping force to hinder the relative movement of the suspension, thereby attenuating the vibration caused by the uneven road surface and improving the ride comfort of the vehicle. The Boost circuit is used to increase the amplitude of the energy-feeding voltage to avoid the voltage amplitude being too small to be used. The dynamics model of the 1 / 4 energy-feedback suspension is established as follows: Among them, m s is the sprung mass; m u is the unsprung mass; z s is the sprung mass displacement; z u is the unsprung mass displacement; z r is the vertical displacement of the road surface; k s is the suspension stiffness; k t is the tire stiffness; c s is the suspension damping coefficient; F is the actuating force; The expression of electromagnetic damping force of linear motor is as follows: Among them, k e is the back electromotive force coefficient of the linear motor; k i is the thrust coefficient of the linear motor; R is the equivalent resistance of the energy feeding circuit; r is the internal resistance of the linear motor; A PI feedforward model predictive control algorithm is designed for the Boost circuit, including a voltage outer loop based on the combination of PI control and feedforward control, and a current inner loop based on model predictive control; at the same time, an inductor current observer and a load disturbance observer are designed in the voltage outer loop and the current inner loop; an inductor current observer is designed for the current inner loop in the PI feedforward model predictive control algorithm of the Boost circuit, and the inductor current observer converts the input voltage u i and the inductor resistor R L Reconstructed into a disturbance variable w, the observed value of the inductor current and the disturbance observation value are used to calculate the inductor current value at the next moment A load disturbance observer is designed for the voltage outer loop in the Boost circuit PI feedforward model predictive control algorithm. The load disturbance observer converts the inductor current observation value estimated by the inductor current observer into As the input of the load disturbance observer, it is used to estimate the output voltage With output current Estimate the load resistance And calculate the reference inductor current i ref (k+1); In the voltage outer loop of the Boost circuit PI feedforward model predictive control algorithm, the reference inductor current is composed of the correction of the output voltage steady-state error by the PI controller and the observer output reference current derived from the power balance principle; in the cost function design of the current inner loop model predictive controller, the errors of the output voltage and the inductor current are considered at the same time; The inductor current observer for the Boost circuit PI feedforward model predictive control algorithm is expressed as follows: Among them, i L (k) is the sampled current at time kT, w(k) is the disturbance variable; is the observed value of the inductor current, is the input voltage u i and the inductor resistor R L The disturbance observation caused by The input voltage u i and the inductor resistor R L Modeled as a disturbance variable w, w(t)=u i (t)-i L (t)R L Among them, u i is the input voltage, R L is the inductor resistance; The load disturbance observer for the Boost circuit PI feedforward model predictive control algorithm is expressed as follows: From the above formula, we can see that The observed value of the inductor current in the inductor current observer is used in To estimate the output voltage With output current And calculate the load resistance The setting of load disturbance observer is used to avoid affecting the stability of energy feeding voltage when the load changes suddenly. The estimation of load resistance by the observer enables the energy feedback control system to respond quickly to disturbances, thereby improving the robustness and dynamic response speed of the energy feedback control system. The model based on PI control and feedforward control is: Among them, i ref is the reference current, u ref is the reference voltage, is the observed value of the disturbance variable, is the observed value of the load equivalent resistance, is the output voltage observation value; The model based on model predictive control is: Where d is the duty cycle, The observed value of the disturbance variable, is the output voltage observation value, i ref is the reference current, is the observed value of the inductor current, L is the inductor value, and T is the sampling period.

2. The energy-feedback suspension control system based on the PI feedforward model predictive control algorithm according to claim 1 is characterized in that: The voltage outer loop uses a PI controller combined with feedforward compensation to calculate the reference current, and the current inner loop outputs the optimal duty cycle through a model predictive control module, driving the pulse width modulator to generate a pulse signal to control the on and off of the MOSFET, so that the energy feeding voltage is stabilized at the desired voltage value.

Images