1. Introduction
An electro-hydraulic proportional multi-way valve is a core control component in hydraulic systems. As the demand for reliability, control accuracy, response speed, and intelligence and integration in hydraulic systems for equipment such as construction machinery and agricultural machinery continues to rise, the electro-hydraulic proportional multi-way valve becomes increasingly critical [
1]. However, the electro-hydraulic proportional multi-way valve exhibits strong nonlinearity, and the application environment is subject to multi-source disturbances, making it difficult to ensure its dynamic and static characteristics, which, in turn, directly affects system performance [
2,
3]. Therefore, there is an urgent need to conduct research on the high-performance dynamic and static control of electro-hydraulic proportional multi-way valves under multi-source disturbance excitation.
To improve the performance of hydraulic valves, scholars at home and abroad have conducted extensive research on the performance of hydraulic systems. Hao Wang et al. established a mathematical model for the main valve displacement and output flow of high-pressure, high-flow-rate, high-frequency water hydraulic proportional flow valves, analyzing the influence of various parameters on the dynamic characteristics of the flow valves [
4]. Xinqiang Liu et al. analyzed the erosion characteristics of servo valve spool orifices and developed an anti-erosion bionic structural optimization design [
5]. Chengzhen Wang et al. established a corresponding mathematical model based on the working principle of a pilot-assisted balancing valve based on the hydraulic–mechanical feedback principle, analyzing the flow control characteristics, overcompensation characteristics, and the influence of backpressure on the main flow [
6]. Fuyan Lyu et al. addressed the issue of significant axial transient vibration on both sides of the valve and minor radial transient vibration caused by the opening of the check valve, analyzing the vibration characteristics of high-pressure, long-distance, thick slurry pipelines using hydraulic active check valves [
7]. Chenghu Guo analyzed the dynamic characteristics involved in the opening, unloading, overflowing, and closing of two-stage safety valves under impact [
8]. Qianpeng Chen et al. addressed the clamping problem of the valve spool caused by thermal deformation, conducted viscous heating experiments on a V-shaped groove valve spool, and analyzed the thermal deformation and clamping force characteristics of a hydraulic valve spool [
9]. Haibo Xie et al. established static equilibrium equations of a pilot-assisted load control valve and derived the mutual influence relationship between backpressure and other parameters, thereby analyzing the backpressure compensation characteristics of the pilot-assisted load control valve [
10]. The aforementioned research on hydraulic valve performance primarily focuses on the influence of hydraulic fluid on the overall valve performance, without considering the influence of controllers.
Domestic and international scholars have conducted a series of studies on the high-performance control of hydraulic valves to improve their performance. Chen Lei et al. proposed state feedback control with integral compensation for the displacement control of servo proportional valve cores, compensating for the nonlinearity and uncertain parameters of hydraulic valves and eliminating the steady-state error of displacement tracking [
11]. Ma et al. proposed a sliding mode variable structure control algorithm for nonlinear problems caused by internal valve leakage and uncertainties in the working environment, meeting the basic control requirements of the system and being insensitive to system parameters [
12]. Qiang Gao proposed an improved PWM controller for the position control of the main valve core in high-speed solenoid valves, with the proposed PWM control reducing the maximum impact acceleration by at least 44.7% and maintaining a total opening time delay of less than 0.3 ms [
13]. Tobia et al. derived a controller design model encapsulating key nonlinearities for the position control of directional control valves and employed a cascade control algorithm, effectively compensating for the valve’s nonlinearity and improving position control accuracy [
14]. Xing Yang et al. proposed a pressure adaptive multi-level voltage and sliding mode control algorithm for water hydraulic high-speed on–off valves, enabling the on–off valves to switch rapidly and maintain dynamic characteristics under varying environmental pressures [
15]. The research on hydraulic valve control methods mentioned above provides certain directional guidance for the high-performance control of the core position of the proportional multi-way valve in this paper.
The proportional multi-way valve spool position control link is mainly composed of three parts, namely, a proportional controller, a proportional reducing pressure valve, and a multi-way valve, which belongs to a typical electro-hydraulic servo nonlinear system, and a lot of research has been carried out on the electro-hydraulic servo position control system by scholars at home and abroad. Deptuła et al. proposed a method for designing hydraulic valves by using multivalued logic trees and weighted factors to determine the structure and the method of the operating parameters of two-stage proportional safety valves, addressing the technical issues of the studied proportional valve series [
16]. Stosiak et al. proposed a different fluid model and hybrid friction model for the impact of mechanical oscillation on hydraulic direction control valves, significantly improving the consistency between simulation and experimental results [
17]. Feng et al. proposed an improved PSO algorithm in view of the shortcomings of the basic particle swarm optimization algorithm on the electro-hydraulic position servo system of robot excavators in terms of convergence speed and the avoidance of local optimal solutions [
18]. Wang et al. proposed a counterstep controller based on a tracking differentiator (TD), which reduces the complexity of the implementation process of the backstep controller, analyzes the performance difference and stability between the alternative signal and the original signal, and ensures tracking accuracy [
19]. Sun et al. proposed a sliding mode control method based on the adaptive approximation law to effectively reduce the jitter and improve the robustness of the system [
20]. For precise position adjustment in hydraulic servo systems, Kaixian Ba et al. proposed a self-disturbance control method based on matrix sensitivity, which effectively reduced nonlinear interference and significantly reduced peak error [
21]. Xiangfei Tao et al. proposed a sliding mode backstep control strategy based on the particle swarm optimization algorithm and a neural network disturbance observer to control the system input through the general approximation characteristics of the observer through neural network interference [
22]. Zhang et al. proposed a composite control method based on sliding mode and self-impulse control to solve the problem that the boom tunneling electro-hydraulic servo system is difficult to control under the conditions of parameter changes and uncertain disturbances, and it improved the tracking accuracy and robustness of the position servo system compared with traditional self-interference control, PID, and sliding mode control [
23]. Zhang et al. developed a direct neural adaptive self-disturbance controller to solve the problems of uncertainty and large external disturbance of the electro-hydraulic servo system of high-order nonlinear systems of engineering mining machinery, and simulation results showed that the direct neural adaptive self-disturbance controller has good tracking accuracy and robustness [
24]. Zhuang et al. proposed a control method that combines sliding mode control and extended state observer control for the position control of electro-hydraulic servo systems with nonlinear mechanical dynamics, which effectively compensates for the influence of internal and external disturbances on control performance [
25]. Yu et al. proposed the closed-loop working principle of an electro-hydraulic load-sensing hydraulic power unit for addressing the mismatch in the pressure and flow rate of hydraulic drive components, derived mathematical models for key components with strong nonlinearity and time-varying parameters, and introduced a matching control strategy based on flow rate feedforward and pressure differential compensation feedback. Simulation and experimental results verified the feasibility of the system and its matching control strategy [
26]. Mou et al. designed an active load-sensing variable displacement drive system that does not rely on mechanical structures for load-sensing adjustment to address issues such as a reduced dynamic response speed and limited load-sensing range. They proposed a dual-loop anti-disturbance control method based on energy efficiency mapping, which improved the system’s position control accuracy [
27]. The above research on the position control method of electro-hydraulic servo systems provides a certain directionality for the high-performance position control of the spool position of proportional multi-way valves with a pilot stage.
To realize the high-performance control of the proportional multi-way valve spool position, this paper takes the proportional multi-way valve spool position control system as the research object. By utilizing the linear extended state observer (LESO) to estimate disturbances, combined with sliding mode control (SMC) for its model insensitivity and strong robustness characteristics, this paper analyzes the influence of controller hardware parameters on the output current, and it proposes a sliding mode control algorithm based on a linear extended state observer (abbreviated as LESO-SMC throughout the following text) to address the impact of nonlinear forces, such as hydrodynamic and friction forces, as well as internal and external multi-source disturbances, on the proportional multi-way valve spool position control performance. A comparative analysis is conducted with the PID control method to verify the effectiveness of the proposed control strategy in enhancing the dynamic and static performance of the proportional multi-way valve spool position control in both the pilot stage and the entire valve.
2. Mathematical Modeling and Influencing Factor Analysis of Proportional Multi-Way Valves
2.1. Working Principle
Figure 1 shows a cross-sectional view of a multi-way valve that supports both manual and electro-hydraulic control. Among them, the electro-hydraulic control adopts two plug-in proportional reducing pressure valves to control the pressure or flow of the multi-way valve’s pilot inlet and pilot outlet, thereby driving the multi-way valve spool to actuate. The yellow, red, orange, blue, and green parts shown in the figure respectively represent the pilot oil, high-pressure oil, oil entering the main valve, oil returning to the oil tank, and A/B port oil.
Figure 2 shows a diagram of the hydraulic principle of a multi-way valve, dashed lines indicate oil routes with lower flow.
The proportional reducing pressure valve, as the direct control object of the proportional spool position control system of the proportional multi-way valve, is controlled by a proportional controller. The proportional controller receives instruction signals applied from the outside and outputs control signals to control the outlet pressure of the proportional reducing pressure valve, ultimately driving the multi-way valve spool to actuate. The performance of the proportional controller directly affects the performance of the proportional multi-way valve spool position control system. To ensure precise control of the multi-way valve spool position, a closed-loop system is required for high-precision control. The structure of the proportional multi-way valve spool position control system is shown in
Figure 3.
First, the controller receives external control command signals, which are processed by the microprocessor chip and output as two control signals to the proportional reducing pressure valve. The detection circuit forms a current internal closed loop, ensuring that the outlet pressure of the proportional reducing pressure valve remains stable. The output pressure of the proportional reducing pressure valve controls the actuator movement of the multi-way valve. The LVDT displacement sensor signal, after being conditioned by the detection circuit, forms a position external closed loop, achieving precise control of the multi-way valve actuator position. Due to the limited driving capability of the microprocessor chip, a driving circuit needs to be added to drive the proportional reducing pressure valve.
2.2. Mathematical Model of Proportional Multi-Way Valve
- 1.
Analysis of the controller output current model and influencing factors
The working process of the PWM-driven reverse connection unloading drive circuit is shown in
Figure 4, where the solid line loop represents the charging process, and the dashed line loop represents the unloading process, with the following simplifications:
(1) In practical applications, the proportional electromagnet is generally operated in the linear region, ignoring the nonlinear factors of the proportional electromagnet, and the proportional electromagnet is equivalent to a series connection of resistance and inductance.
(2) During the charging and unloading processes of the proportional electromagnet, there are also sampling resistor, diode, and MOS on-resistances, as well as internal resistance of the power supply. It is assumed that the total resistance during charging and unloading is equal, both being .
When the charging process of the proportional electromagnet is
, the function of the current of the proportional electromagnet to time is
Then, the average current of the [0, T] period is
In the formula, is the power supply voltage (V); is the equivalent inductance of the proportional solenoid (mH); is the equivalent resistance of the proportional solenoid (); and is the current of the proportional solenoid as a function of time (A). is the current of the proportional solenoid at zero initial time (A); is the circuit time constant; and is a constant. It is assumed that the total resistance during charging and unloading is equal, both being . is the PWM duty cycle, ; T is the PWM period (s).
The average current in the period [KT, (K + 1)T] is
To analyze the impact of the controller hardware parameters on the performance of a pilot-operated proportional valve, the proportional valve’s pilot-stage solenoid drive power supply is given as a 24V DC power signal. The current magnitude under different resistances, different inductances, and different PWM frequencies is analyzed, as well as the variation in the current ripple amplitude under different PWM frequencies and PWM duty cycles. The simulation results are shown in
Figure 5.
As shown in
Figure 5, the average current size of the proportional electromagnet is mainly affected by resistance, with the maximum current steady-state value being
; the current size of the proportional electromagnet is less affected by the PWM frequency and inductance parameters, and it can be ignored. In (a), it can be seen that, when the PWM duty cycle is less than 0.5, the current of the proportional electromagnet is very small and can be ignored; when the PWM duty cycle is greater than 0.5, the current of the proportional electromagnet shows a very good linear relationship with the duty cycle, so the adjustment range of the reverse connection unloading drive circuit is between
. It can also be seen in
Figure 5c,e that the current ripple is mainly affected by the PWM frequency and inductance. The higher the PWM frequency, the smaller the current ripple and the smoother the current waveform. The larger the inductance, the smaller the current ripple. Additionally, the maximum amplitude of the current ripple occurs near a duty cycle of 0.5–0.6. The current ripple is one of the important performance indicators of PWM-driven reverse polarity discharge drive circuits. An excessive current ripple can cause the proportional electromagnet to operate frequently, not only increasing the loss and temperature change of the proportional electromagnet itself but also exacerbating the valve core vibration and mechanical wear of the proportional reducing pressure valve, reducing the service life of the proportional reducing pressure valve, and increasing the control difficulty of the proportional reducing pressure valve.
- 2.
State-space model of pilot-type electro-hydraulic proportional valve
Since the single-plate spool of the proportional multi-port valve has symmetrical left and right structures, the transfer function of spool displacement y to spool displacement x of the proportional reducing pressure valve in the pilot stage is
In the formula, is the flow gain coefficient; is the main spool force area of the multi-way valve (m3); is the natural frequency, ; is the damping ratio, ; is the total volume of the main spool of the multi-way valve (m3); is the viscous damping coefficient (N·s/m); is the total mass of the main spool of the multi-way valve (kg); is the oil volume elastic modulus (Pa); is the inner and outer leakage coefficient (m3/s·Pa); and is the flow/pressure gain coefficient of the proportional reducing pressure valve.
The inherent frequency of the proportional controller and the proportional reducing pressure valve is much higher than that of the proportional valve spool position control system. Simplifying the proportional controller and the proportional reducing pressure valve as a proportional element, that is, ensuring that the displacement of the proportional reducing pressure valve spool is linearly related to the input current of the proportional controller, yields the transfer function of the proportional valve spool position control system:
In the formula, .
,
, and
are the displacement, velocity, and acceleration of the spool of the proportional multi-way valve, respectively. Then, the state equation of the proportional multi-way valve is
In the formula, , , , .
3. Control Algorithms
The proportional multi-way valve spool position control system contains nonlinear factors such as hydrodynamic forces, friction forces, electrical signal interference, external load characteristics, time-varying hydraulic parameters, and proportional controllers, as well as internal disturbances and external disturbances such as superimposed random interference from external signals.
To address the impact of nonlinear hydrodynamic forces and multi-source internal and external disturbances on the position control performance of the spool in a pilot proportional multi-way valve, LESO is employed to estimate the total disturbance of the system. LESO-SMC is established, with its structure shown in
Figure 6, thereby improving the control accuracy and anti-interference performance of the spool position control system in the pilot proportional multi-way valve.
A new total disturbance state
x4 is added to the system of Equation (6), and
. Therefore, the state equation of the pilot electro-hydraulic proportional multi-way valve after expansion is
In the formula, , , , , where is the known input part.
3.1. Differential Tracker
For second-order discrete systems,
Derive a differential tracker based on the fastest synthesis function
:
In the formula, represent the system state; are the control parameters of the fastest synthesis function; and is the system sampling period. is the filter factor. Generally, can be taken as an integer multiple of . Adjusting the speed factor regulates the speed of tracking the input signal.
Four state variables are obtained by cascading four differential trackers, and the differential signal
of the fourth differential tracker is discarded. The differential trackers established are
The specific form of
is
In the formula, is the adjustment speed factor, is the filter factor, and is the system sampling period.
3.2. Expansion State Observer
- 1.
Establishment of an extended state observer
The extended state space equation of the proportional multi-port valve spool position control system is shown in Equation (7), where
is the derivative of
and is bounded, ultimately establishing a linear extended observer:
In the formula, , , , .
In the formula, is the state observer gain. Using the bandwidth parameterization method, with the observer bandwidth determined, all poles are configured at , .
- 2.
Stability Analysis of the Expansion State Observer
The error of the extended state observer is the difference between the system state
and the estimated state
. Therefore, the extended state observer error is
, the differential of the extended state observer error is
, and, accordingly, the error equation of the extended state observer is obtained
Assumption: For the error equation of expression (13),
is bounded, meaning that there exists a number, not less than 0, making
,
. From this, the error state space equation of the extended state observer can be derived:
In the formula, ,.
The characteristic polynomial of the error state space equation
of the expanded state observer is
By choosing
to make
be Hurwitz, for any given symmetric positive definite matrix
, there exists a symmetric positive definite matrix
that satisfies the following Lyapunov equation:
The Lyapunov function that defines the error state space equation of the extended state observer is
In the formula, is the smallest eigenvalue of .
From
, the condition for the convergence of the extended state observer is
If the convergence condition of Equation (20) is satisfied, then the expansion state observer designed in this paper is stable, and the estimation error is bound.
3.3. Sliding Mode Control Law
- 1.
Design of sliding mode control law
An LESO simulation model is built using a fourth-order LESO, with the input being two signals, namely, the control law output quantity and the system output, and the output being the control system estimated state of three proportional multi-way valve spool positions and one total disturbance estimated state. SMC takes the difference between the differential tracker’s output and LESO’s estimated state output, along with LESO’s total disturbance estimated output, as inputs, performs calculations, and outputs the control quantity.
To compensate for the impact of the disturbance state variable
x4 on the position control of the spool in the lead-type electro-hydraulic proportional valve, a sliding mode control law is designed. The displacement reference signal of the proportional multi-way valve spool is defined as
, and the actual errors of the system are defined as
,
, and
. Then, the error equation of the system is
The sliding face is selected as
In the formula, , , and , are constants, taking . as the system bandwidth.
The observation error of the system is
,
,
, and the observation sliding surface is selected as
Taking the derivative of Equation (22) yields
The Lyapunov function is set as
Taking the derivative of Equation (25) yields
Based on the exponential convergence law, the control law
is designed as
In the formula, , .
- 2.
Stability Analysis of Sliding Mode Control System
By substituting Equation (27) into Equation (26), we obtain
In
Section 3.2, it can be seen that the expansion state observer converges; then,
is bounded and sufficiently small, so
, and the sliding mode control law is stable.