Full-active hydraulic suspension energy recovery system based on fuzzy logic programming

Abstract

The invention provides a full-active hydraulic suspension energy recovery system based on fuzzy logic programming, which comprises a full-active hydraulic suspension controller in communication connection with a vehicle body vertical acceleration sensor, a wheel vertical acceleration sensor, a hydraulic system pressure sensor and a battery management system so as to collect sensor signals, the electric hydraulic pump is in communication connection with the full-active hydraulic suspension controller, is connected with the battery and is used for converting mechanical energy of the suspension system into electric energy to be recycled to the battery, and the full-active hydraulic suspension controller plans a generating capacity instruction of energy recycling through a fuzzy logic algorithm based on collected sensor signals. The control system of the electric hydraulic pump receives the generating capacity instruction and adjusts the output current and the generating power of the electric hydraulic pump during power generation according to the instruction, so that the problem of unreasonable generating capacity caused by single parameter control is solved, and an accurate mathematical model and a large amount of experimental debugging are not needed; and the method has strong robustness for nonlinearity and parameter time variation of the suspension system.

Description

Technical Field

[0001] This invention relates to the field of suspension technology, specifically to a fully active hydraulic suspension energy recovery system based on fuzzy logic programming. Background Technology

[0002] Traditional suspension consists of three parts: elastic elements, shock absorbers, and force transmission devices. These three parts respectively play the roles of buffering, damping, and force transmission. This type of suspension is a passive suspension, which is difficult to actively adjust the suspension force according to road conditions. The comfort is generally average. When driving, the vehicle will experience ups and downs due to uneven road surfaces. The energy of these ups and downs comes from the vehicle's driving force. At the same time, this energy of ups and downs is dissipated in the vibration of the elastic elements and shock absorbers and cannot be recovered and reused.

[0003] To address this issue, fully active suspension systems that can actively adjust suspension force have emerged on the market. These systems can adjust suspension force according to road condition information, thereby reducing impact and improving ride comfort. They can also convert the mechanical energy of suspension vibration into electrical energy through built-in electromagnetic devices, achieving energy recovery and providing a new way to improve vehicle energy efficiency and solve energy waste.

[0004] In energy recovery, most existing technologies use linear formulas, such as setting a fixed value based on a single parameter SOC, or simple threshold control. These methods cannot handle complex scenarios with multiple coupled parameters and are difficult to resolve the contradiction between severe vehicle vibration and high road surface recovery potential. To address this issue, a fully automatic suspension control method for energy recovery based on deep reinforcement learning has emerged in the market. For example, Chinese patent CN120116681A discloses such a method. This method trains a neural network based on the PPO algorithm to dynamically generate a suspension control strategy, which can achieve synergistic optimization of energy recovery and suspension performance. However, the optimal hyperparameters provided by this method are often not universal and require extensive experimental debugging, which may lead to training failure or strategy deviation. Summary of the Invention

[0005] The purpose of this invention is to overcome one or more shortcomings in the prior art and provide a fully active hydraulic suspension energy recovery system based on fuzzy logic programming.

[0006] To achieve the above objectives, the technical solution adopted by this invention is a fully active hydraulic suspension energy recovery system based on fuzzy logic programming, comprising: The fully active hydraulic suspension controller communicates with the vehicle vertical acceleration sensor, wheel vertical acceleration sensor, hydraulic system pressure sensor and battery management system to collect sensor signals; An electric hydraulic pump, which communicates with the fully active hydraulic suspension controller and is connected to the battery, is used to convert the mechanical energy of the suspension system into electrical energy and recover it to the battery. The fully active hydraulic suspension controller uses fuzzy logic algorithms to plan the power generation command for energy recovery based on the collected sensor signals. The control system of the electric hydraulic pump receives the power generation command and adjusts the output current and power generation of the electric hydraulic pump according to the command.

[0007] Preferably, the fuzzy logic algorithm includes input variable fuzzification, fuzzy rule reasoning, and defuzzification processing. The input variable fuzzification is used to map the sensor signal into a discrete fuzzy set. The fuzzy rule reasoning is based on a preset rule base, performs logical matching on the fuzzy set, and outputs a fuzzy upper limit for power generation. The defuzzification processing is used to convert the upper limit for power generation into the power generation instruction.

[0008] More preferably, the input variable fuzzification includes the following steps: Step 1. Acquire the sensor signals in real time and remove noise through low-pass filtering to obtain sensor data; Step 2. Convert the sensor data into linguistic variables that can be recognized by fuzzy logic. Define 3 to 5 fuzzy sets and corresponding membership functions for each variable.

[0009] More preferably, the membership function is a triangular or trapezoidal function.

[0010] More preferably, the preset rule base has multiple rules, each of which includes system safety, comfort, and energy recovery efficiency, and the priority of the logical matching is system safety > comfort > energy recovery efficiency.

[0011] More preferably, the logical matching adopts the Mamdani algorithm, the activation strength of each rule is the minimum value of the input membership degree, and the output fuzzy set of all activated rules is aggregated by "taking the largest value" to output a fuzzy upper limit of power generation.

[0012] More preferably, the deblurring process uses the centroid method to convert the fuzzy upper limit of power generation into a precise power generation command, and the calculation formula is as follows: ,in This is the power generation command, in kW. `i` is the index variable for the summation operation, used to sequentially traverse each fuzzy set involved in the calculation. `n` is the total number of fuzzy sets. To output the membership degree of the fuzzy set, This is the center value of the corresponding fuzzy set.

[0013] Preferably, the fuzzy logic algorithm repeats every 5ms and dynamically updates the power generation command based on real-time parameters to ensure adaptation to changes in operating conditions.

[0014] Preferably, the sensor signals include vehicle body vertical acceleration signal, wheel vertical acceleration signal, hydraulic system pressure signal, battery SOC value, and maximum charging current.

[0015] More preferably, the vehicle vertical acceleration sensor is installed at the vehicle's center of gravity to collect the vehicle's vertical acceleration signal, the wheel vertical acceleration sensor is installed on the lower control arm of the suspension to collect the wheel's vertical acceleration signal, and the pressure sensor is deployed at a predetermined node in the suspension hydraulic line to collect the hydraulic system pressure signal.

[0016] Preferably, the power generation command includes maximum current and maximum power, and the control system matches the power generation command by adjusting the AC and DC axis current and power generation power of the electric hydraulic pump.

[0017] More preferably, the operation of the electric hydraulic pump is divided into four quadrants. In the second and fourth quadrants, the electric hydraulic pump is in power generation mode, where hydraulic energy is converted into mechanical energy to drive the motor to rotate. The motor then converts the mechanical energy into electrical energy, thus realizing energy recovery.

[0018] Due to the application of the above technical solution, the present invention has the following advantages compared with the prior art: The fully active hydraulic suspension energy recovery system based on fuzzy logic programming provided by this invention includes a fully active hydraulic suspension controller that is communicatively connected to a vehicle vertical acceleration sensor, wheel vertical acceleration sensors, hydraulic system pressure sensors, and a battery management system to collect sensor signals, and an electric hydraulic pump that is communicatively connected to the fully active hydraulic suspension controller and connected to a battery to convert the mechanical energy of the suspension system into electrical energy for recovery to the battery. By enabling the fully active hydraulic suspension controller to plan the power generation command for energy recovery based on the collected sensor signals using a fuzzy logic algorithm, the control system of the electric hydraulic pump receives the power generation command and adjusts the output current and power generation of the electric hydraulic pump according to the command. This solves the problem of unreasonable power generation caused by single parameter control, eliminates the need for precise mathematical models and extensive experimental debugging, and has strong robustness to the nonlinearity and time-varying parameters of the suspension system. Attached Figure Description

[0019] Figure 1 This is a system architecture diagram of the fully active hydraulic suspension energy recovery system in this invention.

[0020] Figure 2 This is a structural diagram of the electric hydraulic pump and battery in this invention. Detailed Implementation

[0021] The fully active hydraulic suspension energy recovery system based on fuzzy logic programming provided by this invention includes: The fully active hydraulic suspension controller communicates with the vehicle vertical acceleration sensor, wheel vertical acceleration sensor, hydraulic system pressure sensor and battery management system to collect sensor signals; An electric hydraulic pump, which communicates with the fully active hydraulic suspension controller and is connected to the battery, is used to convert the mechanical energy of the suspension system into electrical energy and recover it to the battery. The fully active hydraulic suspension controller uses fuzzy logic algorithms to plan the power generation command for energy recovery based on the collected sensor signals. The control system of the electric hydraulic pump receives the power generation command and adjusts the output current and power generation of the electric hydraulic pump according to the command.

[0022] The fuzzy logic algorithm includes input variable fuzzification, fuzzy rule reasoning, and defuzzification processing.

[0023] Input variable fuzzification is used to map sensor signals into discrete fuzzy sets. Input variable fuzzification includes the following steps: Step 1. Acquire sensor signals in real time and remove noise through low-pass filtering to obtain sensor data; These sensor signals include: Vehicle body vertical acceleration signal (a_body), unit m / s 2 This reflects the intensity of vehicle body vibration and is used to assess comfort requirements. The stronger the vibration, the more power generation needs to be limited to reduce damping. Wheel vertical acceleration signal (a_wheel): unit m / s 2 This reflects the impact intensity between the wheel and the road surface and is used to assess the energy recovery potential; the stronger the impact, the higher the recovery potential. Hydraulic system pressure signal (P_hyd): Unit MPa, reflects the load of the hydraulic circuit, used to assess system safety. The higher the pressure, the less power generation is needed to avoid overload. Battery SOC value (soc): Unit is %, which reflects the current state of charge of the battery. The lower the soc, the higher the acceptable power generation. Maximum charging current of battery (I_max): Unit is A. It reflects the real-time charging capability of the battery. The larger I_max is, the higher the allowable power generation.

[0024] Step 2. Convert the sensor data into linguistic variables that can be recognized by fuzzy logic. Define 3 to 5 fuzzy sets and corresponding membership functions for each variable. The membership functions are triangular or trapezoidal functions.

[0025] Specifically as follows: The fuzzy set defined by the vehicle body acceleration signal (a_body) is: {slight (S), moderate (M), severe (L)}, with a universe of discourse of [0, 1.5, 3] m / s. 2(Based on vehicle comfort threshold calibration), membership function: "small" corresponds to vehicle stability (low comfort requirements, energy can be recovered first), "large" corresponds to severe vibration (requires reduction in power generation to reduce suspension damping). The fuzzy set defined by the wheel vertical acceleration signal (a_wheel) is: {low energy (L), medium energy (M), high energy (H)}, with a universe of discourse of [0,3,6] m / s. 2 Membership function: "large" corresponds to severe road bumps (strong relative motion of the suspension and high recovery potential); The fuzzy set defined by the hydraulic system pressure signal (P_hyd) is: {Safety(S), Warning(M), Danger(H)}, with a universe of discourse of [5,12,20] MPa (based on the rated pressure setting of the hydraulic components). The membership function is: "High" corresponds to approaching the system's pressure limit, requiring limitation of power generation. The fuzzy set defining the battery SOC value (soc) is: {low (L), medium (M), high (H)}, with a universe of discourse of [20%, 50%, 80%], and a membership function: "low" corresponds to the battery being able to accept a large amount of charging, and "high" corresponds to the need to reduce power generation to avoid overcharging; The fuzzy set defining the maximum charging current (I_max) of the battery is: {weak(W), medium(M), strong(S)}, with a universe of discourse of [5,15,25]A. The membership function is: "large" corresponds to a strong battery charging capability (allowing for higher power generation).

[0026] Fuzzy rule-based reasoning, based on a pre-defined rule base, performs logical matching on fuzzy sets and outputs a fuzzy upper limit for power generation. The pre-defined rule base contains multiple rules, each encompassing system safety, comfort, and energy recovery efficiency. These rules take the following forms: The logical matching uses the Mamdani algorithm. The activation strength of each rule is the minimum value of the input membership degree. The output fuzzy set of all activated rules is aggregated by "taking the largest value" to output the fuzzy upper limit of power generation. The priority of logical matching is system safety > comfort > energy recovery efficiency.

[0027] Deblurring is used to convert the upper limit of power generation into a power generation command. Specifically, deblurring uses the centroid method to convert the fuzzy upper limit of power generation into a precise power generation command. The calculation formula is as follows: ,in This is the power generation command, in kW. `i` is the index variable for the summation operation, used to sequentially traverse each fuzzy set involved in the calculation. `n` is the total number of fuzzy sets. To output the membership degree of the fuzzy set, The center value of the corresponding fuzzy set (e.g., "Minimum" = 0.5kW, "Small" = 1.5kW, "Medium" = 3kW, "Large" = 5kW, "Maximum" = 7kW).

[0028] After receiving the power generation command, which includes the maximum current and maximum power, the control system of the electric hydraulic pump matches the power generation command by adjusting the AC and DC axis current and power generation of the electric hydraulic pump. At the same time, the fuzzy logic algorithm repeats every 5ms to dynamically update the power generation command according to the real-time parameters, ensuring that it adapts to changes in working conditions.

[0029] It should be noted that, in order to accurately collect signals, the vehicle vertical acceleration sensor is installed at the vehicle's center of gravity to collect the vehicle's vertical acceleration signal, the wheel vertical acceleration sensor is installed on the lower control arm of the suspension to collect the wheel's vertical acceleration signal, and the pressure sensor is deployed at a predetermined node in the suspension hydraulic line to collect the hydraulic system pressure signal.

[0030] Furthermore, the operation of the electric hydraulic pump is divided into four quadrants. In the second and fourth quadrants, the electric hydraulic pump is in power generation mode, where hydraulic energy is converted into mechanical energy to drive the motor to rotate. The motor then converts the mechanical energy into electrical energy, thus achieving energy recovery.

[0031] like Figure 1 and Figure 2 As shown, the fully active hydraulic suspension energy recovery system provided by this invention is a complex dynamic control system integrating perception, decision-making, and execution. It balances vehicle driving comfort, handling, and stability by adjusting hydraulic damping and support force in real time. Its core components include: Perception layer: A vertical acceleration sensor for the vehicle body, installed at the vehicle's center of gravity, is used to collect real-time acceleration signals (unit: m / s²) in the vertical direction of the vehicle body. 2 This reflects the intensity of vehicle body vibration, such as a sudden increase in acceleration when going over speed bumps, providing core data for comfort assessment; A wheel vertical acceleration sensor, deployed inside the wheel rim, is used to detect the wheel's vertical acceleration relative to the ground (unit: m / s²). 2 This allows us to determine the magnitude of road surface excitation, such as situations where the displacement changes drastically on a bumpy road surface. Hydraulic system pressure sensors integrated into the main hydraulic circuit and various branch pipelines are used to monitor the real-time pressure of hydraulic oil (unit: MPa) to prevent system overpressure (exceeding 20MPa may cause pipeline rupture) and ensure operational safety.

[0032] Decision-making level: The fully active hydraulic suspension controller (ECU) receives sensor data from the sensing layer and calculates the optimal power generation command when the vehicle acceleration exceeds 1.5 m / s². 2 In the event of severe vibration, the fully active hydraulic suspension controller will instruct the hydraulic system to reduce damping force to prevent a deterioration in comfort; when a sharp turn is detected, body roll will be suppressed by adjusting the pressure difference between the two sides of the suspension.

[0033] Execution layer: The electric hydraulic pump, which serves as the power source, is driven by an electric motor to provide high-pressure oil to the hydraulic circuit. The working pressure is 5~20Mpa. The displacement can be adjusted according to the instructions of the fully active hydraulic suspension controller to achieve continuous adjustment of the damping force. The shock absorber has a built-in bidirectional piston. Damping is adjusted by controlling the flow resistance of hydraulic oil in the cylinder. When the piston is driven by the relative movement of the vehicle body and the wheels, the hydraulic oil flows through the throttle orifice through the control valve to generate damping force. Adjusting the opening of the control valve can change the throttle area and change the damping coefficient in real time. Hydraulic pipelines and control valve assemblies include high-pressure oil pipes, proportional relief valves, directional valves, etc., which are responsible for the delivery of hydraulic oil and the distribution of pressure / flow. The proportional relief valve can quickly release excess pressure, while the directional valve controls the flow of hydraulic oil, enabling separate adjustment of the tension / compression damping of the shock absorber.

[0034] Energy and auxiliary layers: The battery provides power to the electric hydraulic pump, such as a 48V, 400V, or 800V power supply. It also serves as an energy storage device for the energy recovery system, storing the recovered electrical energy. The cooling system reduces the hydraulic oil temperature (normal operating temperature 40-80℃) through a radiator and cooling fan, preventing the oil temperature from becoming too high, which would lead to a decrease in viscosity and a reduction in system efficiency.

[0035] The energy recovery function of this fully active hydraulic suspension energy recovery system is achieved through a conversion chain of "vibration mechanical energy → hydraulic energy → electrical energy". The core is to recover and utilize the vibration energy that is wasted as heat in traditional passive suspensions. The specific structure and recovery path are as follows: Step 1: Harnessing mechanical energy from vibration. When a vehicle is in motion, uneven road surfaces, such as potholes or gravel, cause the wheels to bounce up and down. This motion is transmitted through the suspension linkage to the shock absorber piston, which then reciprocates (stroke 50-200mm), converting the relative motion between the wheels and the vehicle body into the mechanical energy of the piston.

[0036] Step 2: Mechanical energy is converted into hydraulic energy. The piston movement forces hydraulic oil to flow inside the shock absorber cylinder, forming high-pressure hydraulic energy (the pressure is proportional to the force on the piston). For example, the piston area is 0.01m². 2 When subjected to a force of 2000N, the hydraulic oil pressure can reach 20MPa.

[0037] Step 3: Hydraulic energy is converted into electrical energy. High-pressure hydraulic oil enters the hydraulic motor through the energy recovery branch, driving the motor rotor to rotate. The speed varies with the flow rate, usually 1000-3000 r / min, which drives the three-phase permanent magnet synchronous motor to cut magnetic field lines and generate electrical energy. At this time, the electric hydraulic pump switches to "driven mode" and no longer consumes electrical energy, but serves as an intermediate link for the transmission of hydraulic energy.

[0038] Step 4: Energy storage and reuse. The electrical energy output by the generator is rectified and regulated, and then charged to the battery by the BMS. The charging current is dynamically adjusted according to the SOC. For example, when the SOC is less than 30%, the maximum current is 25A. The stored electrical energy can be used to drive the electric hydraulic pump, vehicle electrical appliances, etc., reducing the energy consumption of the engine or main battery.

[0039] The normal operation and energy recovery function of this fully active hydraulic suspension energy recovery system are achieved through dynamic switching of the fully active hydraulic suspension controller. During normal driving, the electric hydraulic pump actively outputs high-pressure oil to suppress vehicle vibration by adjusting the damping force of the shock absorbers. At this time, the energy flow is "battery → electric hydraulic pump → hydraulic energy → damping heat loss". During energy recovery, when the relative motion between the vehicle body and the wheels is violent, such as on a bumpy road, and the battery status allows (SOC < 80%), the fully active hydraulic suspension controller shuts off the active drive of the electric hydraulic pump and uses the piston movement to drive the hydraulic oil to flow in the opposite direction, recovering energy through "hydraulic motor → generator". The energy flow is "vibration mechanical energy → hydraulic energy → electrical energy → battery storage". This coordinated mechanism not only ensures the active adjustment capability of the suspension, but also efficiently recovers energy under suitable operating conditions.

[0040] The electric hydraulic pump consists of a three-phase permanent magnet synchronous motor, a control system, and a pump. The control system is connected to the battery via hardware circuitry. When the three-phase permanent magnet synchronous motor operates in electric mode, the suspension is in an active adjustment state, realizing the vehicle comfort adjustment function. The battery voltage enters the control system, converting DC power into AC power. The amplitude of the AC voltage depends on the internal control of the control system. When the voltage generated by the control system is greater than the voltage of the motor's no-load back EMF at the current speed, current flows from the high-voltage control system to the motor, forming the electric current. When the three-phase permanent magnet synchronous motor operates in generator mode, the suspension is in a passive generator mode, realizing the energy recovery function. When the voltage generated by the control system is less than the voltage of the motor's no-load back EMF at the current speed, current flows from the motor to the control system, and after rectification by the control system, it flows to the battery, forming the generator current.

[0041] In the technical architecture of motor control, adjusting the current control strategy can change the motoring and generating states of a three-phase permanent magnet synchronous motor (PMSM). When the PMSM operates as a generator in an electro-hydraulic pump energy recovery scenario, power control needs to be achieved based on the target power output. The core of this control is to adjust the AC and DC axis currents to ensure the output power stably tracks the command value. Using the power output command as input, control is achieved through a three-stage closed loop (power loop → current loop → voltage loop). For example, in the power closed loop, the current command is generated to calculate the actual power output P_actual = U_dc∙I_dc (collected by DC-side voltage and current sensors); the power error ΔP = P_ref - P_actual. The current is processed by the PI regulator, outputting the q-axis current command i_qref (since i_d=0, no d-axis command is needed): i_qref=K_pPΔP+K_iP∫ΔPdt (K_pP and K_iP are power loop PI parameters, which need to be tuned according to dynamic response requirements, such as overshoot ≤5% and settling time ≤50ms); current closed loop, tracking the current command by acquiring the actual dq-axis current i_d and i_q through the current sensor, calculating the current error Δi_d=0-i_d and Δi_q=i_qref-i_q; the error is output by the PI regulator as dq-axis voltage commands u_dref and u_qref to ensure that the current quickly tracks the command; voltage closed loop: the PWM modulation output converts the dq-axis voltage command into a two-phase stationary coordinate system (α-β axis) voltage command through Park inverse transformation; using space vector pulse width modulation (SVPWM) technology, a three-phase PWM signal is generated to drive the inverter, control the stator winding voltage, and finally realize the closed-loop control of current and power.

[0042] The following section, using typical driving scenarios and based on the aforementioned variable definitions, membership functions, and fuzzy rule base, details the calculation and execution process of the power generation of the fully active electro-hydraulic suspension, and verifies the practical application effect of the fuzzy logic programming algorithm.

[0043] Example 1: Bumpy rural road surface.

[0044] Scene characteristics: The road surface has continuous gravel protrusions, the vehicle body vibrates slightly, the battery state of charge is low, and the energy recovery conditions are ideal.

[0045] Vehicle body vertical acceleration sensor: measured value a_body = 0.6 m / s² 2 After low-pass filtering, the velocity stabilizes at 0.5~0.7 m / s. 2 Wheel vertical acceleration sensor: collected value a_wheel = 5.2 m / s² 2 It reflects high-intensity road impact; hydraulic system pressure sensor: main oil circuit pressure Phyd=11Mpa, no fluctuation after sliding average filtering; battery status: SOC=28%, BMS output maximum charging current Imax=24A.

[0046] Vertical acceleration of the vehicle body: a_body = 0.6 m / s² 2 Corresponding to "Slight (S)", membership degree 0.8; wheel vertical acceleration: a_wheel = 5.2 m / s² 2 The corresponding values ​​are: "High Energy (H)" with a membership degree of 0.9; Hydraulic pressure: Phyd=11Mpa, corresponding to "Safety (S)" with a membership degree of 0.8; Battery SOC: 28%, corresponding to "Low (L)" with a membership degree of 0.9; Charging current: 24A, corresponding to "Strong (S)" with a membership degree of 0.9.

[0047] Activation rule R1 (minor vehicle body issues + high wheel energy + hydraulic safety + low SOC + strong charging current), activation intensity is taken as the minimum value of each membership degree, 0.8; rule R1 outputs "maximum (VH)", corresponding to a fuzzy set center value of 7kW; accurate power generation is calculated using the center of gravity method. P_out = 7kW × 0.8 = 5.6kW The control system sent a 5.6kW power generation command to the electric hydraulic pump, which operated at full load. The battery was charged at a current of 24A, and 0.047kWh of electrical energy was recovered within 30 seconds. The vehicle body vibration did not increase, and the acceleration remained at 0.7 m / s². 2 the following.

[0048] Example 2: Urban expressway.

[0049] Scenario characteristics: Occasionally there are manhole covers protruding on the road surface, the vehicle experiences moderate vibration, the battery is at a moderate state of charge, and a balance needs to be struck between comfort and recycling efficiency.

[0050] Vehicle body vertical acceleration sensor: measured value a_body = 1.8 m / s² 2 After filtering, the velocity stabilizes at 1.7~1.9 m / s. 2 Wheel vertical acceleration sensor: collected value a_wheel = 3.5 m / s² 2 Hydraulic system pressure: Phyd=14Mpa (close to the warning threshold); Battery status: SOC=52%, maximum charging current Imax=16A.

[0051] Vehicle acceleration: 1.8 m / s² 2 Corresponding to "Medium (M)", membership degree 0.6; wheel vertical acceleration: 3.5 m / s² 2 The corresponding value is "Medium Energy (M)" with a membership degree of 0.8; the hydraulic pressure is 14MPa, corresponding to "Warning (M)" with a membership degree of 0.5; the battery SOC is 52%, corresponding to "Medium (M)" with a membership degree of 0.9; and the charging current is 16A, corresponding to "Medium (M)" with a membership degree of 0.7.

[0052] Activate rule R5 (medium vehicle body + medium wheel energy + hydraulic safety + medium SOC + medium charging current) with an activation intensity of 0.5; simultaneously activate some conditions of rule R2 with an activation intensity of 0.2.

[0053] Rule R5 outputs "Medium (M)" and rule R2 outputs "Small (S)". After aggregation, the calculation is performed using the centroid method: P_out=((3kW×0.5)+(1.5kW×0.2)) / (0.5+0.2)=2.57kW Execution result The power generation is controlled at 2.6kW, corresponding to a charging current of 13A, the hydraulic pressure is stabilized at 14~15Mpa, and the vehicle's vertical acceleration does not exceed 2.0m / s². 2 Comfort remains unaffected.

[0054] Example 3: Congested urban road surface Scenario characteristics: The road surface is flat but there are frequent starts and stops, the vehicle vibration is weak, the battery is close to full charge, and the recycling demand is low.

[0055] Vertical acceleration of the vehicle body: a_body = 0.3 m / s² 2 Vertical acceleration of the wheel: a_wheel = 1.2 m / s² 2 Hydraulic pressure: Phyd=8Mpa; Battery status: SOC=83%, maximum charging current Imax=8A.

[0056] Vertical acceleration of vehicle body: 0.3 m / s² 2 Corresponding to "Slight (S)", membership degree 0.9; wheel vertical acceleration: 1.2 m / s² 2 The corresponding values ​​are: "Low Energy (L)" with a membership degree of 0.9; hydraulic pressure: 8MPa corresponds to "Safety (S)" with a membership degree of 0.9; battery SOC: 83% corresponds to "High (H)" with a membership degree of 0.8; charging current: 8A corresponds to "Weak (W)" with a membership degree of 0.7.

[0057] Rule R3 (moderate vehicle body + low wheel energy + hydraulic safety + high SOC + weak charging current) is activated, but the actual vertical acceleration of the vehicle body is "slight", and the rule matching degree is 0.7; Since SOC=83%, the implicit rule of "high SOC limit" is activated, with an activation strength of 0.8.

[0058] Both rules output "Minimum (VL)", with a center value of 0.5kW; Precise power generation: P_out = 0.5kW, unaffected by activation intensity, as "extremely small" is a mandatory limit.

[0059] The generator operates at 0.5kW power with a charging current of 4A, and the battery terminal voltage remains stable at 13.8V. The vehicle body experiences almost no vibration, resulting in optimal comfort.

[0060] Example 4: Hydraulic system nearing overload Scenario characteristics: When a vehicle passes over a 10cm high speed bump, the hydraulic system pressure rises sharply, and system safety must be prioritized.

[0061] Vertical acceleration of the vehicle body: a_body = 2.2 m / s² 2 Vertical acceleration of the wheel: a_wheel = 6.0 m / s² 2 Hydraulic pressure: Phyd = 19 MPa; Battery status: SOC = 40%, Imax = 20 A. A hydraulic pressure of 19 MPa corresponds to "Hazard (H)" with a membership degree of 0.9. Activate rule R4 (any input + hydraulic danger → minimal power generation), activation strength 0.9.

[0062] The power generation is forcibly limited to 0.5kW, the hydraulic system pressure drops to 17MPa within 0.5 seconds to avoid pipeline overload, and the vehicle body vibration recovers to 1.0m / s² within 1 second. 2 the following.

[0063] Implementation effect verification: The four scenarios described above cover typical operating conditions ranging from high recovery potential to safety priority. The fuzzy logic algorithm, through rule-based reasoning, achieves a 92% match between power generation output and scenario requirements. Real-vehicle testing shows that the energy recovery efficiency of this fuzzy logic-based fully active hydraulic suspension energy recovery system fluctuates by ≤5% under different road conditions, and the overshoot of vehicle vibration acceleration is ≤10%, proving that the method can achieve multi-objective collaborative optimization.

[0064] The above embodiments are only for illustrating the technical concept and features of the present invention. Their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They should not be used to limit the scope of protection of the present invention. All equivalent changes or modifications made in accordance with the spirit and essence of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A fully active hydraulic suspension energy recovery system based on fuzzy logic programming, comprising: The fully active hydraulic suspension controller communicates with the vehicle vertical acceleration sensor, wheel vertical acceleration sensor, hydraulic system pressure sensor and battery management system to collect sensor signals; An electric hydraulic pump, which communicates with the fully active hydraulic suspension controller and is connected to the battery, is used to convert the mechanical energy of the suspension system into electrical energy and recover it to the battery. Its features are: The fully active hydraulic suspension controller uses fuzzy logic algorithms to plan the power generation command for energy recovery based on the collected sensor signals. The control system of the electric hydraulic pump receives the power generation command and adjusts the output current and power generation of the electric hydraulic pump according to the command.

2. The fully active hydraulic suspension energy recovery system according to claim 1, characterized in that: The fuzzy logic algorithm includes input variable fuzzification, fuzzy rule reasoning, and defuzzification processing. The input variable fuzzification is used to map the sensor signal into a discrete fuzzy set. The fuzzy rule reasoning is based on a preset rule base, performs logical matching on the fuzzy set, and outputs a fuzzy upper limit for power generation. The defuzzification processing is used to convert the upper limit for power generation into the power generation instruction.

3. The fully active hydraulic suspension energy recovery system according to claim 2, characterized in that: The input variable fuzzification includes the following steps: Step 1. Acquire the sensor signals in real time and remove noise through low-pass filtering to obtain sensor data; Step 2. Convert the sensor data into linguistic variables that can be recognized by fuzzy logic. Define 3 to 5 fuzzy sets and corresponding membership functions for each variable.

4. The fully active hydraulic suspension energy recovery system according to claim 3, characterized in that: The membership function is a triangular or trapezoidal function.

5. The fully active hydraulic suspension energy recovery system according to claim 2, characterized in that: The preset rule base contains multiple rules, each of which includes system safety, comfort, and energy recovery efficiency. The priority of the logical matching is system safety > comfort > energy recovery efficiency.

6. The fully active hydraulic suspension energy recovery system according to claim 5, characterized in that: The logical matching uses the Mamdani algorithm. The activation strength of each rule is the minimum value of the input membership degree. The output fuzzy set of all activated rules is aggregated by "taking the largest value" to output the fuzzy upper limit of power generation.

7. The fully active hydraulic suspension energy recovery system according to claim 2, characterized in that: The deblurring process uses the centroid method to transform the fuzzy upper limit of power generation into a precise power generation command. The calculation formula is as follows: , in This is the power generation command, in kW. `i` is the index variable for the summation operation, used to sequentially traverse each fuzzy set involved in the calculation. `n` is the total number of fuzzy sets. To output the membership degree of the fuzzy set, This is the center value of the corresponding fuzzy set.

8. The fully active hydraulic suspension energy recovery system according to claim 1, characterized in that: The fuzzy logic algorithm repeats every 5ms and dynamically updates the power generation command based on real-time parameters to ensure adaptation to changes in operating conditions.

9. The fully active hydraulic suspension energy recovery system according to claim 1, characterized in that: The sensor signals include vehicle vertical acceleration signal, wheel vertical acceleration signal, hydraulic system pressure signal, battery SOC value, and maximum charging current.

10. The fully active hydraulic suspension energy recovery system according to claim 9, characterized in that: The vehicle vertical acceleration sensor is installed at the vehicle's center of gravity to collect the vehicle's vertical acceleration signal. The wheel vertical acceleration sensor is installed on the lower control arm of the suspension to collect the wheel's vertical acceleration signal. The pressure sensor is deployed at a predetermined node in the suspension hydraulic line to collect the hydraulic system pressure signal.

11. The fully active hydraulic suspension energy recovery system according to claim 1, characterized in that: The power generation command includes maximum current and maximum power, and the control system matches the power generation command by adjusting the AC and DC axis current and power generation power of the electric hydraulic pump.

12. The fully active hydraulic suspension energy recovery system according to claim 11, characterized in that: The operation of the electric hydraulic pump is divided into four quadrants. In the second and fourth quadrants, the electric hydraulic pump is in power generation mode, where hydraulic energy is converted into mechanical energy to drive the motor to rotate. The motor then converts the mechanical energy into electrical energy, thus realizing energy recovery.

Images