Digital closed loop controller for flow electric direct drive valve

Abstract

This application relates to the field of industrial automation control and discloses a digital closed-loop controller for a flow-driven electric direct-drive valve. The controller includes a main control module, a signal acquisition circuit, a drive output module, and a current sampling circuit. The main control module acquires the temperature of the valve body and circuit board, calculates the core temperature of the coil based on an equivalent model of a thermal resistance network, and uses this temperature and the target pressure setpoint to obtain a feedforward control current through multi-dimensional lookup tables and linear interpolation to compensate for hysteresis characteristics. Simultaneously, the PID gain parameters are adjusted in real time based on the core temperature of the coil, and a feedback compensation current is calculated to suppress thermal decay. Finally, the feedforward and feedback currents are superimposed and used to control the drive coil via an inner current loop. This invention effectively solves the problems of thermal drift, hysteresis nonlinearity, and model parameter mismatch in a wide temperature range environment without embedding an internal temperature sensor, thus improving the control accuracy of the valve.

Description

Technical Field

[0001] This invention relates to the field of industrial automation control technology, specifically to a digital closed-loop controller for a flow electric direct drive valve. Background Technology

[0002] Electro-driven flow control valves are core actuators in hydraulic and pneumatic control systems. Their working principle involves applying current to an electromagnetic coil via a drive circuit, generating electromagnetic force to displace the valve core, thereby achieving continuous regulation of fluid pressure or flow. These valves are widely used in automotive automatic transmissions, hydraulic systems in construction machinery, and industrial automation process control due to their fast response and compact structure. Ideally, the output pressure of an electro-driven flow control valve is linearly related to the input current. However, in practical engineering applications, its control accuracy is significantly affected by changes in the internal temperature of the coil and the hysteresis characteristics of the electromagnetic material, causing the actual output to often deviate from the expected target.

[0003] At the physical characteristics level, when an electro-electric direct-drive valve operates under prolonged power or in a high-temperature environment, the coil windings experience a significant temperature rise due to copper losses. This temperature increase not only leads to increased coil resistance but also a decrease in the permeability of the magnetic material, causing thermal decay of the electromagnetic force coefficient—a phenomenon known as "thermal drift." This manifests as a significantly lower actual output pressure at high temperatures compared to low temperatures, even with the same drive current command. Furthermore, the inherent hysteresis characteristics of electromagnetic actuators create a non-linear relationship between the valve's input current and output pressure. The characteristic curves during loading and unloading do not coincide, and the width and shape of the hysteresis loop change with temperature. This further complicates maintaining control consistency across a wide temperature range.

[0004] To address the aforementioned issues, existing technologies primarily attempt to solve them through hardware monitoring or algorithmic compensation, but both have limitations. On the hardware side, accurately obtaining the coil core temperature typically requires embedding a temperature sensor inside the coil winding. This not only increases manufacturing complexity and cost but also compromises the integrity of the valve body structure, introducing potential sealing failures and leakage risks in high-pressure applications. Measuring only the valve body surface temperature results in significant lag due to thermal resistance and capacity between the shell and the internal coil, failing to accurately reflect the core heat accumulation state. On the algorithm side, conventional current closed-loop control can only stabilize the current value and cannot compensate for the thermal decay of the electromagnetic force coefficient. Traditional PID controllers are usually tuned based on model parameters at a single reference temperature. When drastic changes in operating temperature alter the system's open-loop gain, fixed control parameters struggle to balance dynamic response and steady-state accuracy, failing to meet high-precision control requirements. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a digital closed-loop controller for flow-driven electric direct-drive valves, which solves the problems of thermal drift, hysteresis nonlinearity, and electromagnetic force thermal attenuation in flow-driven electric direct-drive valves under wide temperature range operating conditions. Without embedding a temperature sensor inside the coil, it improves the control accuracy and response consistency of the valve control system.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a digital closed-loop controller for a flow-driven electric direct-drive valve, comprising a main control module, a signal acquisition circuit, a drive output module, a current sampling circuit, and a non-volatile memory. The main control module integrates a microprocessor, an analog-to-digital converter, and a timer unit. The signal acquisition circuit is connected to the main control module and is used to condition the valve body surface temperature, circuit board temperature, pressure feedback, and external control signals. The drive output module internally includes a bridge circuit composed of power MOSFETs for driving the coil. The current sampling circuit is connected in series in the coil circuit, and the non-volatile memory is used to store control parameters.

[0007] The main control module is configured to execute digital closed-loop control logic: based on the collected data, it uses the equivalent model of thermal resistance network to calculate the core temperature of the coil, which characterizes the internal heat accumulation state of the electric direct drive valve; using the core temperature of the coil and the target pressure setpoint as indexes, it looks up the table to obtain the feedforward control current, and corrects the PID gain according to the core temperature of the coil to calculate the feedback compensation current; finally, it superimposes the feedforward and feedback currents and outputs the drive signal through the drive output module.

[0008] Preferably, to address the issue that sensors cannot directly measure the internal temperature of the coil, the main control module calculates the core temperature of the coil using the following discretization formula based on the principle of thermal resistance networks.

[0009] ;

[0010] ;

[0011] in, Estimate the temperature for the current moment; The valve body surface temperature; The circuit board temperature; This is the thermal damping weighting coefficient, which characterizes the ratio of internal thermal resistance to external thermal resistance in the heat flow path; This is the base temperature calibration bias. It is a smoothing factor; Calculate the temperature of the coil core at the previous moment.

[0012] Preferably, considering the hysteresis nonlinearity of the electromagnetic actuator, the non-volatile memory stores a multi-dimensional discrete hysteresis reference data table. This data table includes loading feature datasets and unloading feature datasets established for multiple different reference temperatures. The main control module determines the hysteresis direction based on the changing trend of the target pressure setpoint: when the target pressure setpoint is greater than or equal to the previous time value, the loading feature dataset is selected; when the target pressure setpoint is less than the previous time value, the unloading feature dataset is selected.

[0013] Preferably, when calculating the feedforward control current, the main control module performs dual linear interpolation: first, it determines the two adjacent reference temperature ranges where the coil core calculation temperature is located, and then performs linear interpolation of the pressure dimension in the corresponding temperature dataset using the target pressure setpoint to obtain the low-temperature end and high-temperature end reference currents. Subsequently, based on the position of the coil core calculation temperature in the reference temperature range, the two reference currents are weighted and calculated.

[0014] Preferably, to address the performance degradation caused by coil impedance changes and electromagnetic force coefficient attenuation at high temperatures, the main control module employs a gain scheduling strategy to adjust the gain parameters of the PID controller in real time based on the following model:

[0015] ;

[0016] in, These are the corrected PID parameter values. For the system at the standard reference temperature The basic PID parameters are as follows: This is the thermal attenuation compensation coefficient, used to compensate for the physical characteristic that high temperatures cause a decrease in open-loop gain.

[0017] Preferably, the main control module is configured with integral separation and limiting logic: when the absolute value of the system error exceeds a preset separation threshold, the integral accumulation in the PID calculation is stopped; when the feedback compensation current exceeds the allowable range, it is clamped to the boundary value. Simultaneously, the main control module has an error dead zone; when the system error is less than the dead zone threshold, the error is forced to zero to suppress static jitter.

[0018] Preferably, the controller has over-temperature protection and chatter function: when the calculated temperature of the coil core exceeds a preset threshold, the maximum allowable operating current is reduced by a linear slope, and a high-frequency chatter signal is superimposed on the target driving current before the output driving signal. The superimposed signal is used as the given target value of the current inner loop, and the power driving circuit is controlled to turn on and off by adjusting the PWM duty cycle.

[0019] A second aspect of the present invention provides a digital closed-loop controller for a flow-driven electric valve, which, based on the above controller, includes the following steps:

[0020] Step 1: The signal acquisition circuit acquires the valve body surface temperature, circuit board temperature, target pressure setpoint, and real-time pressure feedback value. The main control module performs digital filtering on the data.

[0021] Step 2: The main control module calculates the core temperature of the coil based on the thermal resistance network equivalent model and the temperature difference between the valve body surface and the circuit board.

[0022] Step 3: The main control module determines the hysteresis direction based on the change in the target pressure setpoint, and calculates the feedforward control current by referring to a table and interpolating the temperature calculated by the coil core.

[0023] Step 4: The main control module calculates the system error and corrects the PID parameters in real time based on the calculated temperature of the coil core, and calculates the system error to obtain the feedback compensation current;

[0024] Step 5: The main control module superimposes the feedforward control current and the feedback compensation current to generate a PWM signal to control the output module to drive the coil.

[0025] The technical solution provided by this invention achieves sensorless estimation of the core temperature inside the coil by constructing a thermal resistance network model based on the temperature difference between the onboard and valve body. Using this core temperature as an index, the hysteresis model in feedforward control and the PID gain in feedback control are dynamically corrected in two ways. This solution effectively solves the problems of thermal drift, hysteresis nonlinearity, and electromagnetic force thermal attenuation of the flow electric direct drive valve under wide temperature range operating conditions. Without implanting a temperature sensor inside the coil, it improves the control accuracy and response consistency of the valve control system.

[0026] This invention provides a digital closed-loop controller for a flow-driven electric valve. It offers the following advantages:

[0027] 1. This invention achieves accurate estimation of the core temperature inside the coil of an electric direct-drive valve by constructing a thermal resistance network model based on the temperature difference between the valve body surface and the circuit board. This method utilizes the ratio of internal thermal resistance to external thermal resistance in the heat flow path to solve the technical problem of directly monitoring the core temperature of the coil without damaging the valve body structure to implant an internal temperature sensor. This reduces hardware integration costs and avoids the risk of seal failure caused by sensor intrusion, providing a reliable data foundation for subsequent high-precision thermal compensation.

[0028] 2. This invention establishes a multi-dimensional feedforward compensation mechanism that includes temperature and hysteresis direction dimensions. It can dynamically call the corresponding reference current data based on the real-time calculated temperature and pressure change trends of the coil core. By performing dual linear interpolation operations in the temperature and pressure dimensions, it effectively eliminates the inherent nonlinear characteristics of hysteresis of the electromagnetic actuator and the influence of temperature drift on the output accuracy. This ensures that the valve control system can maintain a linear correspondence between the output pressure and the given command under different operating conditions such as low-temperature start-up or high-temperature continuous operation.

[0029] 3. A PID gain scheduling strategy based on core temperature is adopted, which can compensate for the increase in resistance and the attenuation of electromagnetic force coefficient caused by the rise in coil temperature in real time. By introducing a thermal attenuation compensation coefficient to automatically adjust the proportional, integral and derivative parameters, the system can offset the decrease in open-loop gain under high temperature environment, prevent the control system from having response hysteresis or increased steady-state error, and thus ensure the consistency of dynamic response characteristics of the flow electric direct drive valve in a wide temperature range. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the system hardware composition of a digital closed-loop controller for a flow electric direct-drive valve according to the present invention;

[0031] Figure 2 This is the main flowchart of the control method for a digital closed-loop controller for a flow electro-driven valve according to the present invention;

[0032] Figure 3 This invention provides a control algorithm principle and data flow diagram for a flow electro-electric direct-drive valve digital closed-loop controller. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0034] Please see the appendix Figure 1 - Appendix Figure 3 This invention provides a digital closed-loop controller for a flow electric direct-drive valve, including a main control module, a power supply module, a signal input module, a valve temperature acquisition module, an onboard temperature measurement module, a pressure acquisition module, a drive module, a monitoring output module, and a communication module.

[0035] The main control module is the core computing component of the controller. It is electrically connected to the power module, signal input module, valve temperature acquisition module, onboard temperature measurement module, pressure acquisition module, drive module, monitoring output module, and communication module. The main control module is configured to receive various sensor signals and control commands, execute digital signal processing and closed-loop control algorithms, and output drive control signals.

[0036] The power module is used to provide operating voltage for the various circuit units inside the controller. The input terminal of the power module is connected to an external DC power supply, and the output terminal includes a power isolation circuit and a multi-channel voltage conversion circuit to convert the input DC power into the voltage level required by the system, including analog circuit power supply voltage and digital circuit power supply voltage. The power module is equipped with an independent power supply output port for providing excitation power to the externally connected pressure sensor.

[0037] The signal input module is used to receive command signals from the host computer or external control system. The signal input module includes a control signal input channel and an enable signal input channel. The control signal input channel receives 4-20mA current control signals and converts them into analog voltage signals through a current-to-voltage conversion circuit and a signal amplification circuit. The enable signal input channel receives 0-32V voltage signals and shapes them into logic level signals or analog signals through an isolation or level conversion circuit, and transmits them to the digital input port of the main control module.

[0038] The onboard temperature measurement module is integrated on the main circuit board of the controller and is used to detect the temperature data of the controller circuit board and the environment. The onboard temperature measurement module contains a temperature-sensitive element and converts the collected temperature information into an electrical signal and transmits it to the main control module.

[0039] The valve temperature acquisition module is used to connect to an external temperature sensor installed on the valve body of the electric direct drive valve. The input terminal of the valve temperature acquisition module receives the signal from the external temperature sensor, which is then processed by the signal conditioning circuit and transmitted to the main control module.

[0040] The pressure acquisition module is used to connect to the pressure sensor built into the servo valve. The pressure acquisition module receives the feedback signal from the pressure sensor, and after differential amplification, filtering and analog-to-digital conversion, transmits the digital signal representing the actual pressure to the main control module.

[0041] The drive module connects the output of the main control module to the coil load of the electric direct drive valve. The main control module outputs control signals to the drive module. The drive module contains a power amplifier circuit and a constant current drive circuit, which converts the weak electrical signal into a current signal to drive the servo valve.

[0042] The monitoring output module is used to send the system's current operating status to external monitoring equipment. The main control module sends the collected real-time pressure data or internally calculated status values ​​to the monitoring output module. The monitoring output module converts the digital values ​​into a 4-20mA current signal for output.

[0043] The communication module enables digital communication between the main control module and the external host computer. The communication module adopts a serial communication protocol circuit. External devices can read the internal status parameters of the controller through the communication module and configure the control parameters and temperature compensation coefficient.

[0044] Step S10 mainly involves the initialization process after the system is powered on and the real-time acquisition and digital preprocessing of multi-dimensional sensor signals. This step aims to ensure that the control system is in a defined initial state, eliminate physical noise in the transmission process of sensor analog signals, and convert discrete digital sampled values ​​into standard engineering quantities with actual physical meaning, providing a reliable data foundation for subsequent closed-loop control algorithms. Specifically, step S10 includes the following sub-steps:

[0045] S101, Execute system hardware self-test and load operating parameters. After the main control module powers on, it first executes the startup code to perform integrity verification on the internal random access memory (RAM) and program memory (Flash). After the verification passes, the main control module initializes and configures the internal registers, including setting the system clock frequency, configuring the working mode of the general purpose input / output ports (GPIO), initializing the sampling resolution and sampling frequency of the analog-to-digital converter (ADC), and establishing a direct memory access (DMA) channel. In this embodiment, the ADC is configured in a 12-bit or higher precision continuous scan mode, and the conversion result is directly transferred to the memory buffer through DMA to reduce C After the PU load is applied, the main control module reads the pre-stored system calibration parameters from its internal non-volatile memory unit (such as EEPROM or data Flash). The calibration parameters include at least the zero-point calibration value, full-scale calibration value, and filter coefficient of each sensor. At the same time, the main control module detects the status of the enable signal input channel. If a valid enable high-level signal is detected, it enters the closed-loop control preparation state. If no enable signal is detected or a hardware fault flag is detected, the output current of the drive output module is forcibly set to zero, so that the system remains in a safe standby state. The underlying startup timing and register configuration of the microprocessor are well-known technologies to those skilled in the art.

[0046] The main control module uses a timer interrupt to trigger ADC sampling, synchronously reading the analog voltage values ​​of each sensor channel. The data collected includes: the input control signal representing the target pressure setpoint, the pressure feedback signal representing the current actual pressure, the valve temperature signal representing the valve body surface temperature, and the extreme load temperature signal representing the circuit's ambient temperature. Considering electromagnetic interference present in industrial environments and the thermal noise of the analog circuit itself, single sample values ​​often contain random errors. Therefore, the system performs an oversampling mechanism for each channel within a single control cycle (e.g., 1 ms), continuously acquiring multiple sets of raw data (e.g., 8 or 16 times), and averaging them as the raw sample value for that control cycle. .

[0047] S103, Digital Filtering Processing: For the acquired multi-channel raw data, the main control module executes a digital filtering algorithm to filter out power frequency interference and quantization noise. Since different physical quantities have different change characteristics, this step adopts filtering strategies with different response characteristics for pressure signals and temperature signals. Pressure signals are fast variables and need to retain their high-frequency dynamic characteristics, while temperature signals are slow variables with large thermal inertia and are mainly affected by the environment and self-heating, so it is necessary to suppress measurement jitter. Therefore, this embodiment uses a first-order hysteresis filtering algorithm for processing. This algorithm simulates the physical characteristics of a first-order RC low-pass filter in the time domain.

[0048] For any given acquisition channel, let... The original sampled value at time 1 is The previous moment The filtered output value is Then at the current moment Filtered output value Calculate according to the following formula:

[0049] ;

[0050] In the formula:

[0051] These are the filter coefficients, and their values ​​range from [value range missing]. This coefficient determines the weight of the new sampled value in the output result and reflects the update speed of the system.

[0052] For pressure signals and control input signals, to ensure that the system has a response to pressure surges of less than [value missing], Rapid response capability Set to a larger value, the preferred range is ;

[0053] To obtain a stable temperature reference and eliminate measurement jumps for valve body temperature and onboard temperature signals, Set to a smaller value, the preferred range is The filter coefficients It is stored in a non-volatile memory unit and can be adjusted via a communication interface according to the field conditions.

[0054] S104, Engineering Quantity Normalization Conversion: The digital quantity after filtering is a dimensionless value. The main control module needs to convert it into standard physical engineering units (such as MPa, ℃, mA). The main control module performs linear transformation based on the pre-calibrated sensor characteristic curve.

[0055] Let the filtered digital quantity be... The corresponding physical engineering quantity is The digital value corresponding to the zero point determined during the calibration process of this sensor channel is: The digital quantity corresponding to the full scale is The corresponding minimum physical range is The maximum physical range is ,in, and These are the parameters that the system reads and stores in non-volatile memory units during the factory calibration phase by actually loading physical signals; and Based on the known constants specified in the sensor specifications, the normalized calculation formula is as follows:

[0056] ;

[0057] Through the above steps, the main control module converts the acquired raw electrical signals into standardized data required by the system algorithm: the target pressure setpoint. Real-time pressure feedback value Valve body temperature value and onboard temperature value This process essentially achieves digital calibration of the sensor's zero-point deviation and full-scale gain error, replacing the cumbersome process of adjusting with a potentiometer in traditional analog circuits, and ensuring the system's measurement accuracy across the entire range.

[0058] See attached document Figure 2Step S20 involves constructing a thermal field gradient observation model and estimating the coil core temperature. The core purpose of this step is to solve the technical challenge of directly measuring the true core temperature of an electrically driven valve during operation, as the electromagnetic coil is encapsulated inside the valve body and cannot be directly measured by external sensors. The core temperature of the coil directly determines the resistivity of the wire and the permeability of the magnetic material, which are key physical quantities affecting the electromagnetic force output characteristics. If only the valve body surface temperature is used for compensation, the thermal resistance and heat capacity of the heat transfer medium (such as potting compound or metal casing) will lead to significant hysteresis and attenuation in the measurement results. Therefore, this step uses the valve body temperature and the onboard temperature to construct a differential observer and derives the internal core temperature through a thermodynamic model. Specifically, step S20 includes the following sub-steps:

[0059] S201. An equivalent model of the thermal resistance network and gradient variable extraction are established. Based on the principle of thermodynamic heat transfer, the main control module considers the electric direct-drive valve and controller system as a series thermal resistance network. In this network, the energized electromagnetic coil is the main active heat source. The heat generated follows the second law of thermodynamics and is conducted in the direction of decreasing temperature. The heat flow path is mainly divided into two segments: the first segment is conducted from the coil core through the internal insulation material to the valve body metal shell. The thermal resistance corresponding to this path is defined as the internal thermal resistance. The second segment is conducted from the valve body metal shell through the mounting bracket and air convection to the controller circuit board and the surrounding environment. The thermal resistance corresponding to this path is defined as the external thermal resistance. According to the steady-state heat conduction law, the heat flux through the internal thermal resistance is approximately equal to the heat flux through the external thermal resistance under quasi-steady state. This means that the temperature difference between the coil core and the valve body shell (internal temperature difference) is directly proportional to the temperature difference between the valve body shell and the environment (external temperature difference).

[0060] The main control module extracts the valve body temperature value after preprocessing in step S10. and onboard temperature value ,in, The main control module calculates the difference between the two temperatures, using the system's ambient heat sink as a reference temperature point. The difference This refers to the external temperature difference mentioned above. Its magnitude directly reflects the heat flux density of the current system dissipating heat outward, and thus indirectly reflects the degree of heat accumulation inside the system.

[0061] S202 performs a linearized core temperature estimation calculation. Based on the aforementioned thermal resistance ratio, within the temperature range permissible for engineering applications (e.g., -40°C to 125°C), the complex three-dimensional thermal field conduction is simplified into a one-dimensional linear gradient model. The main control module calls the thermal model parameters stored in the internal non-volatile memory and calculates the estimated temperature of the coil core according to the preset observer formula. .

[0062] The calculation formula is as follows:

[0063] ;

[0064] In the formula:

[0065] The current sampling time The estimated equivalent average temperature of the copper wires and iron core inside the coil;

[0066] The current valve body surface temperature is the collected and filtered value.

[0067] The current temperature of the circuit board after being collected and filtered;

[0068] This is the thermal resistance weighting coefficient, which is physically approximately equal to the ratio of internal thermal resistance to external thermal resistance. When the thermal conductivity of the filling material inside the valve body is poor or the external heat dissipation conditions are good, the internal thermal resistance is relatively large, and the value of this coefficient increases accordingly. In this embodiment, this coefficient... The typical value range is 0.5 to 3.0;

[0069] This is the base temperature calibration bias, in °C, used to correct for sensor installation position errors and inherent system deviations.

[0070] S203, Calibration and establishment of model parameters, coefficients in the formula and These are key parameters for ensuring estimation accuracy. They need to be obtained through offline type testing and calibrated, and pre-written into the controller's non-volatile memory. The specific calibration method is as follows:

[0071] Place the electric direct-drive valve in a constant-temperature test chamber and set the chamber temperature to [temperature value missing]. Different constant driving currents are applied at different ambient temperature points (such as low temperature -20°C, normal temperature 25°C, and high temperature 85°C) to achieve thermal equilibrium.

[0072] At this point, a high-precision voltmeter and ammeter are used to measure the voltage across the coil. With current Calculate the real-time resistance of the coil according to Ohm's law. ;

[0073] Subsequently, the resistance-temperature characteristic formula of copper wire was used. The true average temperature of the coil was calculated by reverse calculation. ,in The standard resistance at 20°C The temperature coefficient of resistance of copper (approximately) C);

[0074] Simultaneously record the data collected by the controller at this time. and Using multiple groups Experimental data were fitted using linear regression with the least squares method to find the optimal solution. and This calibration method ensures that the observation model can accurately reflect the thermal characteristics of a specific type of electric direct drive valve.

[0075] S204, Estimating temperature smoothing and output, although the input signal has already undergone preliminary filtering, to further eliminate factors... The computational noise that may be introduced by the coefficient amplification effect is addressed by the main control module in the calculated... A second smoothing process is performed to obtain the final calculated temperature used in the control algorithm. .

[0076] ;

[0077] In the formula This is the smoothing factor for the temperature observer, typically ranging from 0.05 to 0.2. A smaller value results in a stronger filtering effect, but also a greater lag in tracking temperature changes, ultimately affecting the calculated temperature. The temperature index variable is used as a unified index variable and is then fed to the subsequent hysteresis inverse model module (step S30) for interpolation using a lookup table, and to the PID controller module (step S40) for gain parameter adjustment. Through this step, the system successfully maps changes in internal electromagnetic parameters that are difficult to measure directly into temperature variables that can be calculated in real time, laying the physical foundation for achieving high-precision control across the entire temperature range.

[0078] See attached document Figure 2 Step S30 involves calculating the feedforward control quantity of the hysteresis inverse model based on temperature index. This step is designed based on the thermomagnetic decay characteristics of soft magnetic materials. The electromagnetic coil of the electric direct drive valve generates electromagnetic force through the iron core to drive the valve core displacement. However, the permeability of the iron core material (such as DT4 electrical pure iron) is not a constant value. It not only changes nonlinearly with the magnetic field strength (manifested as a hysteresis loop), but also shows a decreasing trend with increasing temperature. Physically, the increase in temperature intensifies the thermal motion of atoms inside the material, interfering with the orientation of magnetic domains along the direction of the magnetic field, resulting in an increase in the excitation current required to generate the same electromagnetic force. In addition, the fit clearance of the mechanical parts and the viscosity of the lubricating oil also change with temperature, changing the motion friction. Therefore, a single current-pressure model cannot accurately correspond to the entire temperature range. This step establishes a multidimensional discretized inverse model with temperature as the index to calculate the accurate feedforward current that can overcome the hysteresis and friction at the current temperature in real time. Specifically, step S30 includes the following sub-steps:

[0079] S301, construct a multi-dimensional discretized hysteresis reference data table. The main control module's non-volatile storage unit contains several pre-set sets of calibrated pressure-current reference data. These data constitute a three-dimensional lookup table with dimensions of: reference temperature, loading direction, and target pressure. First, determine the reference temperature dimension, defining the reference temperature set as... ,in The total number of calibration temperature points, typically selected to cover the industrial-grade operating range. For example, select (Low Temperature Limit) C (normal temperature reference) C (high temperature limit), secondly, determine the pressure node dimension for each reference temperature. The full-scale pressure range from 0% to 90% is divided into... Discrete pressure nodes The value of needs to balance lookup accuracy and storage space, and the preferred range is . .

[0080] For each reference temperature The system stores two datasets of characteristic curves: the loading (boost) characteristic dataset. ,Include A pressure-current data pair is used to describe the pressure rise path, and the unloading (pressure reduction) feature dataset. ,Include These pressure-current data pairs describe the pressure drop path. The dataset was obtained through offline calibration experiments at a constant reference temperature. The control valve pressure is monotonically increased from zero to full scale and then monotonically decreased back to zero, and the drive current value corresponding to each pressure node is recorded.

[0081] S302, determines the target pressure change trend and hysteresis branch selection, the main control module based on the target pressure setpoint The real-time rate of change is used to determine whether the current system is on the rising or falling edge of the hysteresis loop, and a direction indicator variable is defined. Its value selection logic is as follows:

[0082] ;

[0083] In the formula, Prefk−1 is the target pressure setpoint of the previous control cycle, when When the system determines that it is in a loading state, the main control module will call to load the feature dataset. As a basis for subsequent calculations, when Dk=0, the system determines that it is in an unloading state, and the main control module will call the unloading feature dataset. As a foundation, this logic ensures that the control algorithm can accurately match the physical magnetization and demagnetization paths of the electric direct drive valve.

[0084] S303 performs bilinear interpolation calculations. In order to obtain continuous control quantities between a limited number of calibration data points, the main control module performs bilinear interpolation calculations in both the temperature and pressure dimensions.

[0085] The first step is temperature range positioning and boundary clamping. The main control module calculates the temperature based on the output of step S20. In the reference temperature set Searching in the middle, if Then take (Or use low-temperature limit data directly), if Then take Otherwise, select the one that satisfies the condition. The two adjacent temperature nodes.

[0086] The second step is linear interpolation of the pressure dimension, for a selected reference temperature. ) and the determined direction The main control module looks up the target pressure in the corresponding data table. The interval in which it is located, assuming it is in the data table and Two adjacent discrete pressure nodes are and (satisfy The corresponding rated current values ​​are respectively and At this time, the reference current at this reference temperature The calculation is as follows:

[0087] ;

[0088] like If the current exceeds the maximum pressure node in the data table, then directly take the current value corresponding to the maximum node.

[0089] The third step involves linear interpolation along the temperature dimension to obtain the reference current at the low-temperature end. and high-temperature reference current Then, the main control module calculates the current temperature. Weighted interpolation is performed between these two reference values ​​to calculate the final feedforward control current.

[0090] ;

[0091] The feedforward control output is calculated by the main control module. As an open-loop control quantity, it is temporarily stored for subsequent superposition with the feedback control quantity. Through the above steps, the system directly generates the theoretically required drive current based on the current temperature state and pressure target, eliminating the static operating point offset caused by temperature drift. This allows the subsequent PID closed-loop controller to adjust only the small dynamic residual, thereby improving the system's response bandwidth and control stability.

[0092] Step S40 involves calculating the gain-adaptive PID feedback control quantity. This step aims to address the degradation of closed-loop control performance of the electric direct-drive valve in a wide temperature range due to the temperature drift of the controlled object's physical characteristics. According to automatic control principles, the dynamic performance of the closed-loop control system depends on the total open-loop gain of the forward channel. This total gain is equal to the product of the controller gain and the gain of the controlled object (electric direct-drive valve). At the physical level, as the coil temperature increases, the thermal disturbance of the magnetic domains inside the soft magnetic material intensifies, leading to a decrease in permeability and consequently a decrease in the electromagnetic force coefficient (i.e., current-to-force conversion gain). This means that the gain of the controlled object decays with increasing temperature. If the controller keeps its parameters constant at room temperature, the total open-loop gain at high temperatures will decrease, resulting in a narrower system bandwidth and hysteresis. Therefore, this step introduces a gain scheduling strategy based on temperature observations, causing the controller gain to actively increase with increasing temperature to offset the gain decay of the controlled object and maintain a constant total open-loop gain across the entire temperature range. Specifically, step S40 includes the following sub-steps:

[0093] S401, Calculate system control deviation and dead zone handling; the main control module reads the target pressure setpoint after preprocessing in step S10. With real-time pressure feedback value The main control module calculates the difference between the two to obtain the system error at the current moment. To eliminate the interference of quantization noise from the pressure sensor itself and the background noise of the analog circuit on the control system, and to prevent the actuator from performing high-frequency ineffective actions (oscillations) near the steady-state target, the system introduces nonlinear dead-zone logic in the error calculation stage, setting the dead-zone threshold to... The threshold value is typically set to be 1.2 to 1.5 times the peak-to-peak value of the pressure sensor's maximum measurement noise (e.g., 0.3% to 1.5% of full scale). The judgment logic is as follows: If If so, the system is determined to have entered a steady state, and a forced command is executed. like If the original error value is retained, it will be used in subsequent calculations.

[0094] S402, based on the core temperature, the PID gain parameter is corrected in real time. The main control module calls the coil core temperature calculated from the output of step S20. As scheduling variables, the main control module calculates the proportional coefficient at the current temperature based on the gain scheduling model stored in the non-volatile memory. Integral coefficient and differential coefficients The gain scheduling model uses a linear compensation function based on room temperature. The physical basis for this is that, within the allowable operating temperature range of the electric direct-drive valve (-40°C to 125°C), the thermal decay characteristics of the electromagnetic force coefficient approximately conform to a linear law. The calculation formula is as follows:

[0095] ;

[0096] In the formula: subscript represent (Proportion), (Points) or (differential); These are the corrected PID parameter values ​​at the current calculation temperature. For the system at the standard reference temperature (For example, set at 25°C), the optimized basic PID parameters are obtained by tuning using the critical proportional gain method or the decay curve method. The thermal decay compensation coefficient is a dimensionless positive number. This coefficient is determined by adjusting it at the high-temperature limit (e.g., 125°C) so that the system's step response parameters (rise time, overshoot) remain consistent with those at room temperature. In this embodiment, The typical value range is 0.002 to 0.005.

[0097] The main control module uses the corrected parameters , , Based on the current and historical error sequences, discretized incremental PID calculations are performed. Compared to positional PID, the incremental algorithm only calculates the change in the control quantity. Because its calculation does not depend on the historical cumulative sum, the impact on the system is smaller when processor overflow or control mode switching occurs. The incremental feedback control quantity... The calculation formula is as follows:

[0098] S403 performs incremental digital PID calculation:

[0099] ;

[0100] In the formula: The error is the current time. This is the error from the previous sampling time. This represents the error at the upsampling time.

[0101] S404, the synthesis, integral separation, and limiting of feedback control quantities; the main control module synthesizes the feedback control quantity from the previous control cycle. Increment compared to the current calculation The summation is performed to obtain the total feedback control output at the current moment. During the cumulative update process, in order to improve the dynamic adaptability of the system, the main control module executes the following logic;

[0102] First, perform integration and separation processing, and determine the absolute value of the current error. Does it exceed the preset separation threshold? , It is typically set at 10% to 15% of the target pressure value. This indicates that the system is in a large-scale adjustment process. At this time, in order to prevent the integral term from accumulating excessively and causing overshoot, the integral action is logically temporarily disabled (i.e., in the calculation, let...). like Then the integral action is restored to eliminate steady-state error.

[0103] Second, output the safety limit. The calculated... It is not the final drive current, but rather a correction amount for the feedforward current, defining the allowable range of the system's feedback adjustment. For example, set to the total maximum drive current. The basis for this ratio setting is: feedforward control handles the main output (approximately 80%), while feedback control handles precise adjustment (approximately 20%). If the calculated... If the value exceeds this range, it is directly assigned the boundary value. This limiting operation is mathematically equivalent to restricting the infinite accumulation of the integral term, thus achieving anti-windup functionality. The final result is... The signal is then transmitted to the subsequent output synthesis module for superposition with the feedforward signal. Through this step, the system utilizes a temperature-sensing variable gain strategy to effectively counteract the thermal attenuation effect of the electromagnetic actuator, ensuring consistent dynamic and static performance of the controller across the entire temperature range.

[0104] Step S50 involves synthesizing the total drive current, performing safety limiting protection, and converting the control quantity into a physical drive signal output.

[0105] This step is the final stage in translating the control algorithm into actual physical actions. Its purpose is to integrate the feedforward and feedback components obtained from mathematical calculations into a current command capable of driving the electric direct-drive valve coil, while simultaneously ensuring the safety of the hardware under extreme operating conditions through a multi-level protection mechanism. Specifically, step S50 and its auxiliary functions include the following sub-steps:

[0106] S501, Linear superposition and synthesis of control components. The main control module uses the feedforward control current calculated in step S30. The feedback compensation current calculated in step S40 By performing linear superposition, the total target driving current without limiting is obtained. The calculation formula is as follows:

[0107] ;

[0108] In this composition logic, As the primary control component, it provides approximately 80% to 90% of the system's control output based on the inverse model. This is used to balance the valve core spring stiffness, hydraulic pressure, and foundation friction, ensuring the system's rapid response capability. As a correction component, it provides approximately 10% to 20% of the remaining precise adjustment output to eliminate dynamic residuals caused by inaccurate model parameters, sensor temperature drift, and external vibrations. This "feedforward-dominated, feedback-corrected" topology effectively solves the technical contradiction that simple high-gain PID control is prone to causing oscillations.

[0109] The S502 features multi-level current limiting and over-temperature derating protection. To prevent excessive drive current caused by calculation overflow, sensor failure, or extreme environmental temperatures from burning out the coil insulation, the main control module... Perform multi-level amplitude limiting processing.

[0110] Level 1: Global amplitude hard limiting; the system has a preset maximum allowable continuous operating current. This value is determined based on the wire diameter and rated current carrying capacity of the coil's enameled wire (e.g., set to 0mA). Then forced ;like Considering that this embodiment uses a unidirectional proportional electromagnet, forced The second level is dynamic thermal derating protection based on core temperature. The main control module monitors the coil core temperature estimated in step S20 in real time. The system is set with an over-temperature protection threshold. (For example, set at 130°C according to Class F insulation standards) and recovery threshold (For example, set to 120°C), when At that time, the current limit value remains at 0. when When the system enters linear thermal derating mode, the maximum allowable current is... It decreases linearly with increasing temperature, and the calculation formula is as follows:

[0111] ;

[0112] In the formula This is the derating factor (unit: mA / °C), typically ranging from 50 mA / °C to mA / °C. If the calculated value is... If the current is less than the minimum holding current required to keep the valve open, or if the temperature continues to rise above the limit safety temperature (e.g., 150°C), the output will be forcibly shut off. This soft protection mechanism suppresses temperature rise by actively sacrificing some output characteristics, ensuring that critical equipment will not suffer permanent damage under overheating conditions.

[0113] S503, the chatter signal is superimposed and generated by physical drive. Due to the mechanical contact between the valve core and valve sleeve in the electric direct-drive valve, the static friction is much greater than the dynamic friction. When the control signal changes slightly, the static friction will hinder the movement of the valve core, resulting in control dead zones and hysteresis. To address this, the main control module superimposes a high-frequency, low-amplitude chatter signal on the total target drive current, keeping the valve core in a state of slight oscillation, thus transforming the nonlinear static friction into relatively linear dynamic friction. The final physical command is output to the inner current loop. The calculation is as follows:

[0114] The current sampling time The estimated equivalent average temperature of the copper wires and iron core inside the coil is obtained. This represents the current valve body surface temperature after being collected and filtered. The current temperature of the circuit board is the temperature after it has been collected and filtered. This is a thermal damping weighting coefficient, which is physically approximately equal to the ratio of internal thermal resistance to external thermal resistance. When the internal filling material of the valve body has poor thermal conductivity or the external heat dissipation conditions are good, the internal thermal resistance is relatively large, and the value of this coefficient increases accordingly. In this embodiment, this coefficient... The typical value range is 0.5 to 3.0. This is the base temperature calibration bias, in °C, used to correct for the sensor's own installation position error and the system's inherent bias. In the formula: The DC current component after being limited in step S502. The chatter amplitude is typically set to 1.5 times the rated hysteresis width of the electro-hydraulic direct-drive valve (approximately 2% to 5% of full scale). The flutter frequency is required to be higher than the system's control bandwidth (typically >50Hz) and lower than the inherent resonant frequency of the mechanical structure, preferably in the range of Hz to Hz.

[0115] Subsequently, the main control module will As the given target value for the inner current loop, the main control module reads the actual coil current through the sampling resistor and calculates the required terminal voltage duty cycle using an independent high-frequency current PI control algorithm. ,Right now: The timer unit in the main control module is based on The corresponding square wave signal is generated to control the on and off of the power MOSFET in the drive module, thereby achieving high-precision closed-loop tracking of the coil current.

[0116] S504, parameter configuration and non-volatile memory management. To adapt to different specifications of electrically driven valves or different application environments, this controller provides online parameter configuration functionality through the communication module. The main control module internally maintains a parameter mapping table, mapping key variables involved in the algorithm to readable and writable Modbus register addresses. Specific configurable parameters include, but are not limited to:

[0117] Sensor filter coefficients in step S10 Normalized range parameters ;

[0118] Thermal model coefficients in step S20 , ;

[0119] Hysteresis data table node data in step S30 ;

[0120] PID base gain in step S40 and thermal compensation coefficient ;

[0121] Flutter amplitude in step S503 With frequency .

[0122] An external host computer (such as a PLC or debugging software) sends read and write commands to the controller via an RS485 interface using the Modbus RTU communication protocol. When the main control module receives the parameter write command, it first updates the new parameters to the RAM and makes them effective immediately to verify the control effect. When it receives the save command, it writes the entire parameter set in the current RAM to the EEPROM or DataFlash non-volatile storage area. When the system is powered on and reset next time, the main control module will automatically load the latest parameters from the storage area. This parameter management mechanism clearly supports the functional descriptions of "programmable parameter visualization" and "on-site debugging" in the claims, enabling the same controller hardware to be adapted to various valve body models through software configuration, meeting the control requirements of different flow rates and pressure levels without changing the hardware circuit.

[0123] 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 digital closed-loop controller for a flow-driven electric direct-drive valve, characterized in that, include: The main control module integrates a microprocessor, an analog-to-digital converter, and a timer unit; the signal acquisition circuit is connected to the analog-to-digital converter of the main control module and is used to connect and condition the valve body surface temperature sensor, the circuit board temperature sensor, the pressure feedback sensor, and the external control signal source. The drive output module has its control input terminal connected to the output terminal of the timer unit of the main control module, and contains a bridge circuit composed of power MOSFETs for driving the coil of the electric direct drive valve. A current sampling circuit, connected in series in the loop of the coil and connected to the main control module, is used to provide real-time coil current feedback signals; Non-volatile memory, connected to the microprocessor; The microprocessor is configured to execute digital closed-loop control logic: based on data from the signal acquisition circuit, it calculates the core temperature of the coil using a thermal resistance network equivalent model. ; calculate temperature using the coil core Look up the table to obtain the feedforward control current And calculate the temperature based on the coil core. Adjust the PID gain to calculate the feedback compensation current. Finally, the feedforward and feedback currents are superimposed and the drive signal is output through the drive output module.

2. The digital closed-loop controller for a flow electro-electric direct-drive valve according to claim 1, characterized in that, The main control module is configured to calculate the core temperature of the coil using the following formula. ； ； In the formula, The valve body surface temperature. For circuit board temperature, These are the thermal damping weighting coefficients stored in the non-volatile memory. This is the base temperature calibration bias. This is a smoothing factor.

3. The digital closed-loop controller for a flow electric direct-drive valve according to claim 1, characterized in that, The non-volatile memory stores a multi-dimensional discrete hysteresis reference data table, which contains loading feature datasets and unloading feature datasets established for multiple different reference temperatures. The main control module is configured to: set a target pressure value. The changing trend determines the direction of hysteresis; like If so, then select to load the feature dataset; like Then select the unloading feature dataset.

4. A digital closed-loop controller for a flow electro-electric direct-drive valve according to claim 3, characterized in that, The main control module calculates the feedforward control current. Perform bilinear interpolation at the time: First, determine the calculated temperature of the coil core. The two adjacent reference temperature ranges they are located in; Secondly, the target pressure set value is used. Linear interpolation is performed on the corresponding dataset to obtain the low-temperature reference current and the high-temperature reference current. Finally, the temperature was calculated based on the coil core. At the location within the reference temperature range, the low-temperature reference current and the high-temperature reference current are weighted and calculated.

5. A digital closed-loop controller for a flow electro-electric direct-drive valve according to claim 1, characterized in that, The main control module is also configured to perform PID error dead zone processing and set a dead zone threshold. When the absolute value of the system error between the target pressure setpoint and the real-time pressure feedback value is less than or equal to the dead zone threshold, the error value participating in the PID calculation is forced to be zero.

6. A digital closed-loop controller for a flow electro-electric direct-drive valve according to claim 1, characterized in that, The main control module is configured to correct the PID gain based on the following gain scheduling model: ； In the formula, These are the corrected PID parameter values. These are the basic PID parameters stored in the non-volatile memory. This is the thermal attenuation compensation coefficient, used to compensate for the decrease in open-loop gain caused by the thermal attenuation of the electromagnetic force coefficient at high temperatures.

7. A digital closed-loop controller for a flow electro-electric direct-drive valve according to claim 6, characterized in that, The main control module is also configured to control the feedback compensation current. Perform integral separation and amplitude limiting: When the absolute value of the system error exceeds the preset separation threshold, the accumulation of the integral term in the PID calculation is stopped; When the calculated feedback compensation current When the value exceeds the preset allowable range, it will be clamped at the boundary value.

8. A digital closed-loop controller for a flow electro-electric direct-drive valve according to claim 1, characterized in that, The main control module is also configured to perform over-temperature protection and flutter control: When the core temperature of the coil is monitored When the preset threshold is exceeded, the maximum allowable operating current is reduced by a linear slope. Before the output drive signal, a high-frequency dithering signal is superimposed on the target drive current, and the superimposed signal is used as the given target value of the current inner loop to control the on and off of the drive output module.

9. A digital closed-loop controller for a flow electro-electric direct-drive valve according to claim 1, characterized in that, Also includes: The communication module, connected to the main control module, is used to establish a data communication link between the controller and external devices. The main control module receives external commands through the communication module to configure the thermal damping weighting coefficient, hysteresis reference data table, and PID base gain stored in the non-volatile memory online.

10. A control method for a flow electro-electric direct-drive valve digital closed-loop controller according to claim 1, wherein the flow electro-electric direct-drive valve digital closed-loop controller according to any one of claims 1-9 is characterized in that, Includes the following steps: step S10: The signal acquisition circuit acquires the valve body surface temperature, circuit board temperature, target pressure setpoint, and real-time pressure feedback value; the main control module filters the data. Step S20: The main control module uses the equivalent model of the thermal resistance network to calculate the core temperature of the coil. ; Step S30: The main control module determines the hysteresis direction and calculates the temperature based on the coil core. Look up the table and interpolate to calculate the feedforward control current. ; Step S40: The main control module calculates the temperature based on the coil core. Adjust PID parameters and calculate feedback compensation current. ; Step S50: The main control module superimposes the feedforward and feedback currents to generate a PWM signal to control the drive output module to drive the coil.

Images