Forklift steering state following control system and method based on MPC

Abstract

The invention discloses a forklift steering state following control system and method based on MPC, and belongs to the field of heavy electric forklifts.The forklift steering state following control system comprises a steering manual control unit which is configured to collect a target steering angle instruction of a driver and provide torque feedback; the MPC controller is in communication connection with the steering manual control unit so as to receive the target steering angle instruction and solve an optimal control increment sequence in a future limited time domain through rolling optimization based on a preset forklift steering system dynamics model, a current system state and a system constraint condition; and the steering execution mechanism is electrically connected with the MPC controller so as to receive the control instruction output by the MPC controller at the current moment and drive a rear steering wheel of the forklift to rotate along with the target turning angle instruction. The robustness and safety of the system under the limiting working condition are improved, the energy efficiency of the whole vehicle is optimized, and steering control is improved from passive error correction to active trajectory prediction and optimization control.

Description

Technical Field

[0001] This invention relates to the field of heavy-duty electric forklift technology, and in particular to a forklift steering state following control system and method based on MPC. Background Technology

[0002] As core equipment in logistics warehousing and industrial material handling, the performance of the steering system in heavy-duty electric forklifts directly determines the overall operating efficiency and driving safety of the vehicle. To meet the steering torque requirements under heavy-duty conditions, most existing heavy-duty forklifts use hydraulic power steering systems. However, this system has inherent drawbacks such as the risk of oil leakage, high energy consumption, and slow response.

[0003] With the development of electronic control technology, steer-by-wire systems are gradually being applied to various types of vehicles. By eliminating the mechanical connection between the steering wheel and the steering wheels, sensors are used to collect the driver's intentions, and the steering wheels are driven by a motor or electro-hydraulic actuator, making it possible to achieve more flexible steering control strategies.

[0004] In the field of control for commercial vehicles such as semi-trailers and heavy-duty trucks, various methods based on modern control theory have been proposed to improve their stability under high-speed or extreme operating conditions. For example, Chinese invention patent application CN116620307A discloses a longitudinal and lateral coordinated control method for intelligent semi-trailers during high-speed steering. This method establishes a nonlinear dynamic model of the semi-trailer and designs a nonlinear model predictive controller to coordinately control the front wheel steering angle of the tractor and the braking torque of each wheel, aiming to suppress the folding and tail-wagging phenomena between the tractor and the trailer, thereby improving the lateral stability of the entire vehicle. This scheme demonstrates the advantages of model predictive control in handling multi-constraint optimization problems by imposing constraints on the control variables and their increments.

[0005] However, the aforementioned existing technologies mainly focus on the stability control of vehicles under high-speed turning conditions, and their control objective is to make the overall vehicle state more stable and avoid instability.

[0006] For specific industrial vehicles like heavy-duty forklifts, the core issue in actual operation is often not stability, but rather the accuracy of following steering commands. Specifically, because heavy-duty forklifts often use rear-wheel steering and operate under varied conditions (from high-speed travel under no-load to low-speed micro-motion under heavy load), their steering systems exhibit significant nonlinearity and time lag characteristics. This makes it difficult for traditional control methods or simple angle-correspondence algorithms to guarantee precise and rapid matching between the steering wheel or fingertip controller angle and the actual steering wheel angle under all operating conditions. This easily leads to problems such as overshoot and response delay, affecting operational accuracy and driving comfort, and even creating safety hazards.

[0007] Furthermore, existing technologies do not provide clear guidance on how to deeply couple model predictive control with forklift-specific electro-hydraulic steering actuators to achieve precise control of the steering process.

[0008] Therefore, how to provide a control scheme that can accurately and quickly follow the driver's steering intentions, taking into account the nonlinear and time-delay characteristics of heavy-duty forklift steering systems, is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0009] The technical problem to be solved by this invention is how to provide a control scheme that can follow the driver's steering intention with high precision and fast response, taking into account the nonlinear and time-delay characteristics of the steering system of heavy forklifts.

[0010] In a first aspect, to solve the above-mentioned technical problems, a forklift steering state following control system based on MPC is provided, applied to a heavy-duty electric forklift, comprising: The steering hand control unit is configured to acquire the driver's target steering angle command and provide torque feedback; The MPC controller is communicatively connected to the steering manual control unit to receive the target steering angle command and, based on a preset forklift steering system dynamics model, the current system state, and system constraints, solves for the optimal control increment sequence within a finite time domain through rolling optimization; and The steering actuator is electrically connected to the MPC controller to receive the control command output by the MPC controller at the current moment and drive the rear steering wheel of the forklift to rotate in accordance with the target turning angle command.

[0011] Furthermore, the steering actuator includes: Oil supply unit; A steering ratio control valve assembly is connected to the oil supply unit, and the steering ratio control valve assembly includes a first steering valve and a second steering valve arranged in parallel. The steering cylinder is a double-acting cylinder, with its two chambers independently connected to the first steering valve and the second steering valve, respectively, so as to supply oil to the two chambers independently; The corner detection unit is communicatively connected to the MPC controller to feed back the detected actual corner to the MPC controller.

[0012] Furthermore, the MPC controller is configured as follows: The target steering angle command, the estimated steering angle of the rear steering wheel, the estimated angular velocity of the rear steering wheel, and the control current of the first steering valve and the second steering valve are used as the system state vector; Based on the system state vector and the dynamic model of the forklift steering system, predict the system output at N future sampling times; Construct and solve for the optimal control increment sequence in the next M control time domains that minimizes the cost function, where M≤N, and the cost function includes at least a penalty term for steering follow error; Extract the first element from the optimal control increment sequence obtained from the solution. Combined with the control quantity of the previous moment Generate the actual control quantity at the current moment. .

[0013] Furthermore, in the process of constructing and solving the cost function, the system constraints include at least one or more of the following: the maximum opening degree constraint of the first steering valve and the second steering valve, the stroke constraint of the steering cylinder, and the maximum turning angle constraint of the rear steering wheel.

[0014] Furthermore, the actual control quantity at the current moment This includes a first independent opening command assigned to the first steering valve and a second independent opening command assigned to the second steering valve, to independently control the flow and direction of hydraulic oil entering the two chambers of the steering cylinder.

[0015] Furthermore, the dynamic model of the forklift steering system is established based on the key parameters of the forklift steering system obtained through experimental identification. These key parameters include at least the equivalent moment of inertia and the equivalent damping coefficient.

[0016] Furthermore, the MPC controller is also configured to: The state observer estimates the state vector of the system under continuous operating conditions at the current moment based on the dynamic model of the forklift steering system and the actual steering angle fed back by the steering angle detection unit; wherein the state vector includes at least the estimated steering angle and estimated angular velocity of the rear steering wheel.

[0017] Furthermore, the steering hand control unit includes a fingertip steering controller and a torque feedback device that are coupled to each other, wherein the fingertip steering controller is disposed on one side of the forklift seat.

[0018] Furthermore, the MPC controller is also communicatively connected to the torque feedback unit and is configured to generate torque optimization instructions based on the solved optimal control increment sequence and send them to the torque feedback unit to optimize the driver's torque output.

[0019] In a second aspect, the present invention provides an MPC-based forklift steering state following control method applied to the system, comprising the following steps: S1. Collect the driver's target steering angle command through the steering hand control unit; S2. The actual steering angle of the rear steering wheel is collected in real time through the steering angle detection unit; S3, the MPC controller receives the target turning angle command and the actual turning angle, and estimates the system's current state vector by combining the preset forklift steering system dynamics model; S4. The MPC controller, based on the target turning angle command and the estimated state vector, uses a prediction model to predict the future system output, and under the condition of satisfying the preset system constraints, solves the optimal control increment sequence in the future M control time domains that minimizes the optimization objective function. S5. The first element in the optimal control increment sequence obtained from S4 is... Control quantity compared to the previous moment Superimpose to generate the actual control quantity at the current moment. ; wherein, the actual control quantity This includes independent opening commands for driving the first and second steering valves respectively, so as to independently control the flow and direction of hydraulic oil entering the two chambers of the steering cylinder; S6. The actual control quantity The command is sent to the steering actuator to drive the rear steering wheels to follow the target steering angle. S7. At the next sampling time, repeat steps S2 to S6 to form rolling optimization control.

[0020] Furthermore: In S3, the current state vector of the estimation system includes at least the estimated steering angle and estimated angular velocity of the rear steering wheel; In S4, the prediction model is a dynamic model of the forklift steering system obtained based on experimental identification.

[0021] Compared with the prior art, the embodiments of the present invention have the following beneficial effects: This invention constructs a closed-loop control scheme for steering state following in heavy-duty forklifts by deeply coupling the MPC controller with a hardware system including a fingertip steering controller and a dual-valve independent oil supply electro-hydraulic actuator. This scheme utilizes the rolling optimization and feedback correction mechanism of the MPC to predict and optimize future control sequences online based on the system dynamics model, effectively overcoming the inherent nonlinearity and time-delay characteristics of heavy-duty forklift steering systems. It achieves real-time and precise matching between fingertip controller commands and rear wheel angles, eliminating response delay and overshoot. Simultaneously, by explicitly incorporating multiple system constraints such as valve opening and cylinder stroke into the optimization objectives, and employing dual-valve independent oil supply at the hardware level to accurately execute the fine commands output by the MPC, the system's robustness and safety under extreme conditions are improved, and the overall vehicle energy efficiency is optimized. This transforms steering control from passive error correction to active trajectory prediction and optimization control. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a schematic diagram of the system disclosed in this invention; Figure 2 This is a flowchart of the method disclosed in this invention.

[0024] In the picture: 100. Steering hand control unit; 110. Finger steering controller; 120. Torque feedback device; 200. MPC controller; 311. Steering servo motor; 312. Silent pump; 320. Steering proportional control valve group; 321. First steering valve; 322. Second steering valve; 330. Steering cylinder; 341, 342. Steering angle detection unit; 350. Rear steering wheel. Detailed Implementation

[0025] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] This invention addresses the problems of nonlinear characteristics and time lags, such as tire deformation and steering clearance, and the resulting overshoot or response delay, in heavy-duty electric forklifts, especially rear-wheel steering forklifts, during steering. It proposes a forklift steering state following control system and method based on model predictive control (MPC). This scheme achieves high-precision and fast-response following of the driver's steering intentions by constructing a closed-loop system with a specific hardware architecture and control algorithm.

[0027] The first aspect of this invention discloses a forklift steering state following control system based on MPC, which is applied to heavy-duty forklifts. The following description is in conjunction with the attached... Figure 1 The system will be described in detail.

[0028] Please see Figure 1The system mainly includes a steering hand control unit 100, an MPC controller 200, and a steering actuator. The steering actuator includes an oil supply unit, a steering proportion control valve group 320, a steering cylinder 330, and steering angle detection units 341 and 342. The steering proportion control valve group 320 is connected to the oil supply unit and includes a first steering valve 321 and a second steering valve 322 arranged in parallel. The steering cylinder 330 is a double-acting cylinder, with its two chambers independently connected to the first steering valve 321 and the second steering valve 322, respectively, enabling independent oil supply to the two chambers. The steering angle detection units 341 and 342 are communicatively connected to the MPC controller 200 and are used to feed back the detected actual steering angle to the MPC controller 200.

[0029] Specifically: The steering hand control unit 100 is the interface through which the driver interacts with the system. In a preferred embodiment, the steering hand control unit 100 consists of a fingertip steering controller 110 and a torque feedback unit 120 coupled together. The fingertip steering controller 110 can be a small joystick or knob, and for ergonomic purposes and ease of driver operation, it is typically located on the left or right armrest of the forklift seat. The driver issues a target steering angle command by manipulating the fingertip steering controller 110, representing the angle the driver expects the rear steering wheels 350 to turn. The torque feedback unit 120, connected to the fingertip steering controller 110, applies a feedback torque to the driver's hand based on the vehicle's real-time status and commands from the MPC controller 200. This simulates the road feel of a traditional steering system or provides driver feedback under specific conditions, thereby optimizing the driving experience and operational safety.

[0030] The MPC controller 200 is the decision-making core of the system; it is an electronic control unit based on a microprocessor and embedded software. The input of the MPC controller 200 is communicatively connected to the steering hand control unit 100 to receive the target steering angle signal from the fingertip steering controller 110 in real time. The output of the MPC controller 200 is electrically connected to the control terminals of the steering actuators (i.e., the steering servo motor 311 and the steering proportional control valve group 320). Furthermore, the MPC controller 200 is also connected to the angle detection units 341 and 342 to receive the actual steering angle feedback from the rear steering wheel 350, thus forming a closed-loop control system. The MPC controller 200 has a pre-set dynamic model of the forklift steering system, which describes the dynamic relationship between system inputs (such as valve-controlled current) and outputs (such as wheel angle). Based on this model, the current system state (including the estimated steering angle and angular velocity of the rear steering wheel 350), and pre-set system physical constraints, the MPC controller 200 solves for the optimal control sequence within a finite future time domain through online rolling optimization.

[0031] The steering actuator is a physical device that receives commands from the MPC controller 200 and drives the rear steering wheel 350 to rotate. In this embodiment, the steering actuator is an electro-hydraulic drive-by-wire system, the specific configuration of which is shown below: The oil supply unit consists of a steering servo motor 311 and a silent pump 312 driven by the motor. The steering servo motor 311 receives instructions from the MPC controller 200 and precisely controls the start, stop and speed of the silent pump 312 to provide pressurized oil to the entire hydraulic system on demand, which helps to reduce system energy consumption and noise.

[0032] The steering proportion control valve assembly 320 is connected to the oil outlet of the oil supply unit. The steering proportion control valve assembly 320 is not a simple monolithic valve; it internally contains a first steering valve 321 and a second steering valve 322 arranged in parallel. Both valves are high-response proportional control valves, capable of continuously and precisely adjusting the valve opening according to the input electrical signal.

[0033] The steering cylinder 330 is a double-acting hydraulic cylinder, with its piston rod connected to the steering mechanism of the rear steering wheel 350. The two working chambers of the steering cylinder 330 (i.e., the rod-side chamber and the rodless chamber) are connected to the first steering valve 321 and the second steering valve 322 via independent oil circuits. This unique dual-valve independent oil supply structure allows the system to independently control the flow and direction of hydraulic oil entering the two chambers of the cylinder. For example, when a right turn is required, the MPC controller 200 can control the first steering valve 321 to supply oil to one chamber at a certain opening, while simultaneously controlling the second steering valve 322 to return oil from the other chamber at a different opening. This independent control method offers greater flexibility and control precision than traditional single-valve control, enabling smoother and faster precise displacement of the piston rod within the cylinder.

[0034] The angle detection units 341 and 342 are used to monitor the actual steering angle of the rear steering wheel 350 in real time. To improve the reliability and measurement accuracy of the system, this embodiment adopts a redundant design, including a left-side angle sensor and a right-side angle sensor. These two sensors can be potentiometer-type, magnetoelectric, or photoelectric angle sensors, respectively installed at different positions on the steering knuckle or steering axle of the rear steering wheel 350, for independently detecting the actual steering angle of the rear steering wheel 350. The angle signals they detect are fed back to the MPC controller 200 in real time for status observation and closed-loop control.

[0035] All the above components are tightly connected through signal lines and hydraulic lines to form a complete control system. In this embodiment, the three-dimensional state vector consisting of the forklift rear wheel steering angle, fingertip torque value, and control valve current is used as the input to the MPC controller 200. The MPC controller 200 first uses its internal state observer, combined with the forklift steering system dynamics model, to optimally estimate the state vector of the system under continuous operating conditions at the current moment. Subsequently, the MPC controller 200 uses the state space model to predict the state evolution of the system at the next moment and in the future finite time domain. Finally, the predicted control sequence is rolled through a preset cost function to select a set of control quantities that optimize the performance indicators in the prediction time domain, and applies them to the steering actuator composed of the steering servo motor 311, the first steering valve 321, and the second steering valve 322, driving the rear steering wheel 350 of the forklift to rotate in accordance with the target steering angle command.

[0036] A second aspect of the present invention discloses a forklift steering state following control method based on MPC, which can be applied to the control system described above. The following description, in conjunction with the appendix, further details this method. Figure 2 The method will be explained in detail.

[0037] Please see Figure 2 The method mainly includes the following steps: Step S1: System parameter calibration and model identification Before implementing real-time control, it is necessary to first identify and obtain the key physical parameters of the heavy-duty forklift steering system through experiments, thereby establishing a mathematical model that can relatively accurately reflect the dynamic characteristics of the system. These key physical parameters include, but are not limited to, the equivalent moment of inertia referred to the steering wheels. Equivalent damping coefficient This includes parameters such as dry friction in the steering system and the bulk modulus of the hydraulic fluid. For example, specific excitation signals, such as step signals or sinusoidal sweep signals, can be applied to the steering system, and the system's valve spool displacement, cylinder pressure, and wheel angle responses can be recorded. Then, system identification algorithms (such as the least squares method) can be used to estimate these parameters. The final state-space model can be represented as:

[0038]

[0039] In the formula, Let be the system state vector. To control the input vector, For the system output vector, and It is a nonlinear function.

[0040] Step S2: Signal Acquisition and Preprocessing During forklift operation, the fingertip steering controller 110 in the steering hand control unit 100 collects the target steering angle signal input by the driver in real time. Meanwhile, the actual steering angle of the rear steering wheel 350 is collected in real time through redundantly configured steering angle detection units 341 and 342. To improve signal quality, the acquired raw signal can be low-pass filtered to remove high-frequency noise. Additionally, the actual turning angle can be adjusted. The actual angular velocity of the rear steering wheel at 35° is obtained by differential measurement or by state observation. .

[0041] Step S3: State Observation and Estimation The target angle is obtained from the signal acquired and processed in step S2. Actual turning angle and / or actual angular velocity The input is fed into the state observer built into the MPC controller 200. The state observer, combined with the steering system dynamics model established in step S1, estimates the full-dimensional state vector of the system under continuous operating conditions at the current moment. Since state variables such as internal pressure of the hydraulic cylinder and valve core displacement may be difficult to measure directly, or the measurement cost may be too high, the introduction of a state observer allows these variables to be accurately estimated. This estimates the state vector. It should include at least the estimated steering angle of the rear steering wheels at 350 degrees. With estimated angular velocity These are the foundation for model prediction.

[0042] Step S4: MPC Trajectory Planning and Optimization The driver's target turning angle command The current system state estimated in step S3 These inputs are fed into the optimization solver inside the MPC controller 200. The controller predicts the system output for N future sampling times based on a discretized prediction model. This discretized prediction model can be expressed as:

[0043]

[0044] In the formula, Indicates in Always Prediction of the state at any given moment; Indicates in Decision-making at all times Constantly control input; This represents the predicted system output; , , It is the discretized system matrix.

[0045] The controller constructs an optimization objective function. This function aims to minimize a set of performance metrics over the prediction time domain. A typical cost function takes the following form:

[0046] In the formula, To predict the time domain, The desired output trajectory, i.e., the target turning angle. Its derivative is zero. To control the increment, The weighting matrix is ​​used to adjust the penalties for following error, control severity, and terminal error, respectively. This represents the weighted Euclidean norm.

[0047] During the optimization process, the preset system constraints must be met. These constraints reflect the physical limits of the system, ensuring that the solved control commands are practically executable and safe. Typical constraints include:

[0048]

[0049]

[0050] For example, and This can represent the maximum and minimum opening of the valve ports of the first directional valve 321 and the second directional valve 322; This represents the maximum physical turning angle of the rear steering wheel at 350 degrees to prevent interference between the tire and the frame.

[0051] Under the above constraints, the solver searches for a solution that makes the objective function... Minimize the optimal control increment sequence over the next M control time domains ,in To control the time domain, and .

[0052] Step S5: Output control command The first element in the optimal control increment sequence obtained in step S4. Extract and combine with the actual control quantity at the previous sampling time. The actual control quantity that needs to be applied to the actuator at the current moment is calculated:

[0053] This strategy of applying only the first control variable in the sequence embodies the principle of MPC rolling optimization, which can promptly correct control actions after obtaining new feedback information.

[0054] Step S6: Drive steering execution The actual control quantity calculated in step S5 Send to the steering actuator. Specifically, The signal is first sent to the driver of the steering servo motor 311, which instructs it to drive the silent pump 312 to operate at a specific speed, providing the system with the required pressure and flow. Meanwhile, This includes independent opening commands assigned to the first steering valve 321 and the second steering valve 322, respectively. These two independent commands are converted into corresponding current signals, driving the valve cores of the two valves to move, thereby independently and precisely controlling the flow rate and switching direction of hydraulic oil entering the two chambers of the steering cylinder 330. The resulting cylinder displacement drives the rear steering wheel 350 to rotate through the steering mechanism, achieving a steering action consistent with the driver's target steering angle command. In addition, the MPC controller 200 can also generate a torque optimization command that matches the current operating condition based on the solved optimal control sequence and send it to the torque feedback unit 120, enabling the driver to perceive the force state of the wheels or the constraint limits of the system, thus optimizing human-machine interaction.

[0055] Step S7, Scrolling Optimization When the next sampling moment arrives, the system jumps back to step S2, re-acquires the latest signal, and repeats the entire process from step S2 to step S6. This cycle repeats continuously, forming a rolling optimization control of the steering process, enabling the system to adapt to changing operating conditions and disturbances in real time and always maintain optimal following performance.

[0056] In the above method, the MPC controller 200 estimates the state vector through the state observer. At least include the estimated steering angle of the rear steering wheels. With estimated angular velocity The prediction model used is based on the dynamic model of the forklift steering system obtained from experimental identification, which ensures that the model can accurately reflect the real physical characteristics of the controlled object.

[0057] In summary, this invention creatively combines the MPC algorithm with targeted electro-hydraulic drive-by-wire hardware, successfully solving the steering state following problem of heavy-duty forklifts under all working conditions, especially under heavy load and low speed conditions, achieving precise "point-and-go" control, and significantly improving work efficiency and safety.

[0058] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A forklift steering state following control system based on MPC, applied to heavy-duty electric forklifts, characterized in that, include: The steering hand control unit (100) is configured to acquire the driver's target steering angle command and provide torque feedback; The MPC controller (200) is communicatively connected to the steering hand control unit (100) to receive the target steering angle command and, based on the preset forklift steering system dynamics model, the current system state and system constraints, solves the optimal control increment sequence in the future finite time domain through rolling optimization. as well as The steering actuator is electrically connected to the MPC controller (200) to receive the control command output by the MPC controller (200) at the current moment and drive the rear steering wheel (350) of the forklift to rotate in accordance with the target turning angle command.

2. The forklift steering state following control system based on MPC according to claim 1, characterized in that, The steering actuator includes: Oil supply unit; A steering ratio control valve assembly (320) is connected to the oil supply unit. The steering ratio control valve assembly (320) includes a first steering valve (321) and a second steering valve (322) arranged in parallel. The steering cylinder (330) is a double-acting cylinder, and its two chambers are independently connected to the first steering valve (321) and the second steering valve (322) respectively, so as to supply oil to the two chambers independently; The corner detection unit (341, 342) is communicatively connected to the MPC controller (200) to feed back the detected actual corner to the MPC controller (200).

3. The forklift steering state following control system based on MPC according to claim 2, characterized in that, The MPC controller (200) is configured to: The target steering angle command, the estimated steering angle of the rear steering wheel, the estimated angular velocity of the rear steering wheel, and the control current of the first steering valve (321) and the second steering valve (322) are used as the system state vector; Based on the system state vector and the dynamic model of the forklift steering system, predict the system output at N future sampling times; Construct and solve for the optimal control increment sequence in the next M control time domains that minimizes the cost function, where M≤N, and the cost function includes at least a penalty term for steering follow error; Extract the first element from the optimal control increment sequence obtained from the solution. Combined with the control quantity of the previous moment Generate the actual control quantity at the current moment. .

4. The forklift steering state following control system based on MPC according to claim 3, characterized in that, In the process of constructing and solving the cost function, the system constraints include at least one or more of the following: the maximum opening degree constraint of the valve ports of the first steering valve (321) and the second steering valve (322), the stroke constraint of the steering cylinder (330), and the maximum turning angle constraint of the rear steering wheel (350).

5. The forklift steering state following control system based on MPC according to claim 3, characterized in that, The actual control quantity at the current moment This includes a first independent opening command assigned to the first steering valve (321) and a second independent opening command assigned to the second steering valve (322) to independently control the flow and direction of hydraulic oil entering the two chambers of the steering cylinder (330).

6. The forklift steering state following control system based on MPC according to claim 1, characterized in that, The dynamic model of the forklift steering system is established based on the key parameters of the forklift steering system obtained through experimental identification. These key parameters include at least the equivalent moment of inertia and the equivalent damping coefficient.

7. The forklift steering state following control system based on MPC according to claim 1, characterized in that, The MPC controller (200) is also configured to: The state observer estimates the state vector of the system under continuous operating conditions at the current moment based on the dynamic model of the forklift steering system and the actual turning angle fed back by the turning angle detection unit (341, 342); wherein the state vector includes at least the estimated turning angle and estimated angular velocity of the rear steering wheel (350).

8. The forklift steering state following control system based on MPC according to claim 3, characterized in that, The steering hand control unit (100) includes a fingertip steering controller (110) and a torque feedback device (120) that are coupled to each other, wherein the fingertip steering controller (110) is located on one side of the forklift seat.

9. The forklift steering state following control system based on MPC according to claim 8, characterized in that, The MPC controller (200) is also communicatively connected to the torque feedback unit (120) and is configured to generate torque optimization instructions based on the solved optimal control increment sequence and send them to the torque feedback unit (120) to optimize the driver's torque output.

10. A forklift steering state following control method based on MPC applied to the system described in any one of claims 1-9, characterized in that, Includes the following steps: S1. The driver's target steering angle command is collected through the steering hand control unit (100); S2. The actual steering angle of the rear steering wheel (350) is collected in real time by the steering angle detection unit (341, 342); S3, MPC controller (200) receives the target turning angle command and the actual turning angle, and estimates the current state vector of the system by combining the preset forklift steering system dynamics model; S4. The MPC controller (200) predicts the future system output based on the target turning angle command and the estimated state vector using a prediction model, and solves for the optimal control increment sequence in the future M control time domains that minimizes the optimization objective function under the preset system constraints. S5. The first element in the optimal control increment sequence obtained from S4 is... Control quantity compared to the previous moment Superimpose to generate the actual control quantity at the current moment. ; wherein, the actual control quantity This includes independent opening commands that drive the first steering valve (321) and the second steering valve (322) respectively, so as to independently control the flow and direction of hydraulic oil entering the two chambers of the steering cylinder (330); S6. The actual control quantity The command is sent to the steering actuator to drive the rear steering wheels (350) to follow the target steering angle. S7. At the next sampling time, repeat steps S2 to S6 to form rolling optimization control.

11. The forklift steering state following control method based on MPC according to claim 10, characterized in that: In S3, the state vector of the estimation system at the current moment includes at least the estimated rotation angle and estimated angular velocity of the rear steering wheel (350); In S4, the prediction model is a dynamic model of the forklift steering system obtained based on experimental identification.

Images