Intelligent numerical control hydraulic system and control method

Abstract

This invention provides an intelligent CNC hydraulic system and control method, including a sensing and monitoring unit comprising a multi-source heterogeneous sensor network deployed at key nodes throughout the system. The invention is rationally designed, using a multi-source heterogeneous sensor network deployed throughout the system to simultaneously collect multi-dimensional operating parameters such as pressure, flow rate, temperature, and vibration. Edge computing is used to complete data preprocessing and fault-sensitive feature extraction. Relying on an SVM-LSTM dual-model fusion fault intelligent diagnosis module and integrating a fault tree FTA analysis submodule, it achieves accurate classification of fault types, rapid location of fault points, and root cause tracing. It also provides early warning of early system faults, significantly improving fault diagnosis accuracy compared to traditional threshold alarm methods. Furthermore, it can identify fault types that traditional methods cannot detect, such as directional valve sticking, predicting fault development trends in advance and effectively preventing equipment damage caused by fault escalation.

Description

Technical Field

[0001] This invention mainly relates to the field of hydraulic transmission, specifically to an intelligent numerical control hydraulic system and its control method. Background Technology

[0002] Hydraulic transmission technology, with its significant advantages such as high power density, strong load resistance, wide speed range, and large output force and torque, has been widely used in high-end equipment manufacturing and industrial production fields such as CNC machine tools, engineering machinery, metallurgical equipment, aerospace, mining machinery, and injection molding. It is one of the core basic technologies for transmission and control of modern industrial equipment. Existing CNC hydraulic systems mostly adopt the core control architecture of electro-hydraulic proportional valves, servo valves + PLC controllers. They realize the digital regulation of hydraulic system pressure and flow through electro-hydraulic control components, and complete the CNC operation under fixed working conditions in conjunction with the preset PLC control program.

[0003] During the operation of specific embodiments, the inventors discovered the following defects: Hydraulic system failures are characterized by their high degree of concealment, complex causes, and rapid spread. Early-stage failures such as internal leakage, valve core wear, oil contamination, seal aging, and pump cavitation are difficult to identify through conventional monitoring methods. Most existing CNC hydraulic systems only have basic fault alarm functions such as pressure over-limit and overload shutdown, providing only post-fault alarms and failing to predict, accurately locate, or assess early-stage faults. Furthermore, existing systems lack intelligent fault diagnosis algorithms based on multi-sensor data fusion and remaining service life prediction models, making it impossible to digitally manage the system's health status throughout its entire lifecycle. This easily leads to unplanned equipment downtime, severely impacting the continuity of industrial production. Moreover, equipment maintenance relies on manual inspection and experience-based judgment by technicians, resulting in low efficiency and high costs, and compromising the long-term reliability and safety of the system.

[0004] It should be noted that the above content falls within the scope of the inventor's technical knowledge. Due to the vast and complex nature of the technical content in this field, the above content of this application does not necessarily constitute prior art. Summary of the Invention

[0005] 1. The technical problem that the invention aims to solve: This invention provides an intelligent numerical control hydraulic system and control method to solve the technical problems existing in the background art.

[0006] 2. Technical Solution: To achieve the above objectives, the technical solution provided by the present invention is as follows: an intelligent numerical control hydraulic system and control method, comprising a hydraulic power unit, a hydraulic execution unit, an electro-hydraulic control valve group, a sensing and monitoring unit, and a central control unit. The hydraulic power unit is connected to the hydraulic execution unit via the electro-hydraulic control valve group, the sensing and monitoring unit is communicatively connected to the central control unit, and the central control unit is electrically connected to the hydraulic power unit and the electro-hydraulic control valve group respectively. The sensing and monitoring unit includes a multi-source heterogeneous sensor network, which is deployed at key nodes throughout the system to synchronously collect multi-dimensional operating parameters such as pressure, flow rate, temperature, vibration, displacement, and oil contamination. The central control unit integrates an edge computing diagnostic node, an SVM-LSTM dual-model fusion intelligent fault diagnosis module, a full lifecycle health management module, a digital twin virtual mapping module, and a collaborative control execution module; The edge computing diagnostic node is used for data preprocessing, feature extraction, and local anomaly alarm. The SVM-LSTM dual-model fusion fault intelligent diagnosis module includes an SVM sub-module and an LSTM sub-module, which are used for fault classification, localization and early warning. The full life cycle health management module is used for health status assessment, remaining life prediction, and generation of hierarchical operation and maintenance decisions. The digital twin virtual mapping module is used for fault scenario reproduction and virtual optimization of control strategies; The collaborative control execution module is used for adaptive adjustment of system control parameters and hierarchical safety protection.

[0007] Furthermore, the multi-source heterogeneous sensor network includes pressure sensors, flow sensors, temperature sensors, vibration sensors, displacement sensors, and oil contamination sensors; The pressure sensor and flow sensor are respectively installed at the pump port of the hydraulic power unit, the inlet and outlet of each valve of the electro-hydraulic control valve group, and each working oil chamber of the hydraulic actuator. The vibration sensors are installed in the pump body of the hydraulic power unit, the valve body of the electro-hydraulic control valve group, and the end of the cylinder body of the hydraulic actuator unit. The displacement sensor is installed at the valve core of the electro-hydraulic control valve group and the piston rod of the hydraulic actuator; The temperature sensor and the oil contamination sensor are installed in the system oil tank and the main return oil pipeline.

[0008] Furthermore, the edge computing diagnostic node incorporates an adaptive variational Bayesian filtering algorithm and a multi-scale feature fusion algorithm. The edge computing diagnostic node is also equipped with a local storage unit for caching system operation data and supporting encrypted transmission.

[0009] Furthermore, the SVM-LSTM dual-model fusion fault intelligent diagnosis module also has a built-in fault tree FTA analysis submodule, which is preset with four types of typical faults: system vibration and abnormal noise, pressure abnormality, flow instability, and component failure.

[0010] Furthermore, the full lifecycle health management module divides the system health status into four levels: healthy, sub-healthy, minor fault, and serious fault.

[0011] Furthermore, the digital twin virtual mapping module constructs a high-precision digital twin with all elements, including the system's geometric model, physical model, and behavioral model.

[0012] Furthermore, the hydraulic power unit includes an intelligent variable displacement piston pump with a built-in electronic control unit (ECU), an accumulator, and an oil temperature regulation unit. The ECU of the intelligent variable displacement plunger pump communicates bidirectionally with the collaborative control execution module; The oil temperature regulating unit is connected to the central control unit.

[0013] Furthermore, the electro-hydraulic control valve assembly includes an intelligent electronic proportional valve and a multi-position multi-way solenoid directional valve; The intelligent electronic proportional valve is equipped with a valve core displacement sensor and a drive control unit. The multi-position multi-way electromagnetic directional valve is equipped with a valve position feedback sensor.

[0014] Furthermore, the hydraulic actuator includes a chuck clamping hydraulic cylinder, a tool post indexing hydraulic cylinder, a tailstock sleeve hydraulic cylinder, and a spindle shifting hydraulic cylinder, all compatible with the CNC lathe.

[0015] A control method for an intelligent CNC hydraulic system, wherein the intelligent CNC hydraulic system comprises the following steps: S1: End-to-end data synchronous acquisition: Multi-dimensional time-domain operating parameters of key nodes in the hydraulic system are synchronously acquired through a multi-source heterogeneous sensor network and transmitted to the edge computing diagnostic node. S2: Edge data preprocessing and feature extraction. The edge computing diagnostic node performs filtering and noise reduction preprocessing on the raw data, extracts fault-sensitive feature indicators and completes local anomaly alarms, and then transmits the processed data to the dual-model fusion fault intelligent diagnosis module and the digital twin virtual mapping module respectively. S3: Dual-model fusion for intelligent fault diagnosis and tracing. It completes fault classification and location through the SVM sub-module, completes root cause tracing of faults by combining the FTA sub-module, and completes early fault warning and trend prediction through the LSTM sub-module. The results are synchronously transmitted to the corresponding modules. S4: Full lifecycle health management and operation and maintenance decision generation. Based on fault diagnosis results and component performance degradation models, it assesses the system health status, predicts the remaining service life, and generates corresponding hierarchical operation and maintenance strategies. S5: The digital twin virtual mapping module's digital twin synchronous mapping and simulation optimization uses real-time data to drive the synchronous mapping between the digital twin and the physical system, reproduce fault scenarios, and simulate and verify optimization solutions. The optimal solution is then output to the collaborative control execution module. S6: Adaptive Cooperative Control and Hierarchical Safety Protection. The cooperative control execution module dynamically adjusts the system control parameters based on the fault diagnosis results and the optimal control scheme, and executes hierarchical safety protection actions according to the severity level of the fault.

[0016] 3. Beneficial effects: Compared with the prior art, the technical solution provided by this invention has the following advantages: This application utilizes a multi-source heterogeneous sensor network deployed throughout the entire system chain to simultaneously collect multi-dimensional operating parameters such as pressure, flow rate, temperature, and vibration. It combines edge computing to complete data preprocessing and fault-sensitive feature extraction. Relying on an SVM-LSTM dual-model fusion intelligent fault diagnosis module and integrating a fault tree FTA analysis submodule, it achieves accurate classification of fault types, rapid location of fault points, and root cause tracing. It also provides early warning of early system faults, significantly improving fault diagnosis accuracy compared to traditional threshold alarm methods. Furthermore, it can identify fault types that traditional methods cannot detect, such as directional valve sticking, predict fault development trends in advance, and effectively prevent equipment damage caused by fault escalation. The full lifecycle health management module can accurately assess the health status of each system component and quantitatively predict its remaining service life based on fault diagnosis data, historical operating data, and hydraulic component performance degradation models. It also generates tiered maintenance strategies based on health levels, including routine inspections, key monitoring, downtime maintenance, and emergency shutdowns. Furthermore, it can upload maintenance data to the MES system and industrial internet maintenance platform via the OPCUA protocol, achieving digitalization and standardization of maintenance processes. This solves the problems of traditional hydraulic systems relying on manual experience for troubleshooting, slow fault location, and poor targeted maintenance, significantly shortening fault diagnosis time, reducing maintenance labor costs and unplanned equipment downtime, and extending the mean time between failures (MTBF) of the equipment. Detailed Implementation

[0017] To facilitate understanding of the present invention, a more comprehensive description of the invention will be given below with reference to relevant embodiments, and several embodiments of the invention will be provided. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that the disclosure of the invention will be more thorough and complete.

[0018] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "page," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicating orientation or positional relationships, are based on the indicated orientation or positional relationships and are only for the convenience of describing this invention and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention.

[0019] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0020] In this invention, unless otherwise explicitly specified and limited, the terms "installed," "connected," "linked," "fixed," "provided with," and "located in" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0021] Example An intelligent numerical control hydraulic system and control method are provided. To achieve the above objectives, the technical solution provided by the present invention is as follows: An intelligent numerical control hydraulic system and control method are provided, comprising a hydraulic power unit, a hydraulic actuator unit, an electro-hydraulic control valve group, a sensing and monitoring unit, and a central control unit. The hydraulic power unit is connected to the hydraulic actuator unit via the electro-hydraulic control valve group. The sensing and monitoring unit is communicatively connected to the central control unit. The central control unit is electrically connected to the hydraulic power unit and the electro-hydraulic control valve group, respectively. The architecture of hydraulic power unit, hydraulic actuator unit, electro-hydraulic control valve group, sensing and monitoring unit and central control unit has built a complete closed-loop management system of power output-control regulation-action execution-state perception-intelligent decision-making within the hydraulic system. It replaces the traditional open-loop / semi-open-loop extensive architecture, realizes precise distribution and control of hydraulic oil pressure, opens up the real-time transmission channel of state data, and completes the millisecond-level response from intelligent decision-making to action execution. It fundamentally solves the defects of traditional system control and monitoring disconnect and power and load mismatch, and realizes full-link controllability, measurability and optimizability of hydraulic system. The sensing and monitoring unit includes a multi-source heterogeneous sensor network, which is deployed at key nodes throughout the system. It is used to synchronously collect multi-dimensional operating parameters such as pressure, flow, temperature, vibration, displacement, and oil contamination. It can accurately identify early hidden faults such as pipeline leakage, valve abnormality, and component wear, and control the core fault cause of hydraulic oil contamination from the source. It provides a comprehensive and accurate data foundation for intelligent diagnosis and control optimization, and solves the pain points of insufficient monitoring dimensions and difficulty in capturing early fault characteristics in traditional systems. The central control unit integrates an edge computing diagnostic node, an SVM-LSTM dual-model fusion fault intelligent diagnosis module, a full life cycle health management module, a digital twin virtual mapping module, and a collaborative control execution module. The central control unit locally constructs a full-process intelligent management and control system that integrates data processing, fault diagnosis, health management, simulation optimization, and execution control in the hydraulic system, replacing the traditional single program control function of PLC. It can complete local intelligent management and control without relying on a host computer. The five modules form a complete intelligent closed loop of data processing, diagnostic analysis, management decision-making, simulation optimization, and execution feedback within the system, solving the core defects of traditional systems such as low intelligence level and disconnect between control and diagnosis. The edge computing diagnostic node is used for data preprocessing, feature extraction, and local anomaly alarm. It primarily enables real-time front-end processing of collected data locally within the hydraulic system. Through built-in algorithms, it performs signal noise reduction, fault-sensitive feature extraction, and local anomaly alarm, effectively suppressing signal distortion caused by electromagnetic interference, significantly shortening the data processing and anomaly response cycle, and preventing fault escalation. The accompanying local storage unit can cache operational data, providing data support for fault tracing and model optimization. It constructs the first line of defense for data processing within the system, ensuring the accuracy of subsequent diagnostic analysis data. The SVM-LSTM dual-model fusion fault intelligent diagnosis module includes an SVM sub-module and an LSTM sub-module, which are used for fault classification, localization and early warning. The full life cycle health management module is used for health status assessment, remaining life prediction, and generation of hierarchical operation and maintenance decisions. The digital twin virtual mapping module is used for fault scenario reproduction and virtual optimization of control strategies; The collaborative control execution module is used for adaptive adjustment of system control parameters and hierarchical safety protection. As the core of intelligent decision execution, this module realizes collaborative adaptive control and hierarchical safety protection of multiple components and multiple loops within the hydraulic system. It can dynamically adjust the pump displacement and valve opening to achieve precise matching of pressure, flow rate and load requirements, significantly improving the stability and anti-interference capability of the system under varying operating conditions. It can realize the coordinated control of the action sequence of the four execution loops of the CNC lathe, improving the operating efficiency and safety of the equipment. It implements hierarchical safety protection for different fault levels, constructing a full-level safety protection system and forming closed-loop control, further improving the system control accuracy and operational stability.

[0022] Furthermore, the multi-source heterogeneous sensor network includes pressure sensors, flow sensors, temperature sensors, vibration sensors, displacement sensors, and oil contamination sensors; The pressure sensor and flow sensor are respectively installed at the pump port of the hydraulic power unit, the inlet and outlet of each valve of the electro-hydraulic control valve group, and each working oil chamber of the hydraulic actuator. The vibration sensors are installed in the pump body of the hydraulic power unit, the valve body of the electro-hydraulic control valve group, and the end of the cylinder body of the hydraulic actuator unit. The displacement sensor is installed at the valve core of the electro-hydraulic control valve group and the piston rod of the hydraulic actuator; The temperature sensor and the oil contamination sensor are installed in the system oil tank and the main return oil pipeline.

[0023] Furthermore, the edge computing diagnostic node incorporates an adaptive variational Bayesian filtering algorithm and a multi-scale feature fusion algorithm to suppress electromagnetic interference noise in the acquired signals. Simultaneously, it extracts three types of fault-sensitive feature indicators—fluctuation entropy, frequency domain energy ratio, and wavelet packet energy entropy—from the pressure, flow, and vibration time-series signals, respectively. The edge computing diagnostic node is also equipped with a local storage unit for caching system operation data and supporting encrypted transmission.

[0024] Furthermore, the SVM-LSTM dual-model fusion intelligent fault diagnosis module also incorporates a fault tree FTA analysis submodule. This submodule is pre-defined with four typical fault categories: system vibration and abnormal noise, pressure anomalies, flow instability, and component failure. The SVM-LSTM dual-model fusion intelligent fault diagnosis module achieves precise fault classification and location, as well as early warning within the hydraulic system. The SVM submodule achieves a 96.7% accuracy rate in diagnosing high-frequency system faults, quickly pinpointing fault locations and significantly reducing the difficulty of troubleshooting. The LSTM submodule provides early warning of faults 10-15 minutes in advance, preventing component damage and unplanned downtime caused by fault escalation. Combined with the FTA fault tree analysis submodule, it enables root cause tracing of faults, upgrading the hydraulic system from passive maintenance to proactive prediction of faults.

[0025] Furthermore, the full lifecycle health management module classifies the system health status into four levels: healthy, sub-healthy, minor fault, and severe fault. This module constructs a full lifecycle health management system for components within the hydraulic system. It can quantitatively assess the health status of core components and complete the four-level health level classification. Differentiated management strategies are generated for different levels. By dynamically adjusting parameters to compensate for component performance degradation, unnecessary downtime and replacement are avoided. It can quantitatively predict the remaining service life of components and generate planned operation and maintenance plans, replacing the traditional periodic replacement mode and reducing operation and maintenance consumable costs. At the same time, it generates hierarchical operation and maintenance strategies, realizing predictive proactive operation and maintenance, which significantly improves the long-term operational reliability of the system and reduces the full lifecycle operation and maintenance costs.

[0026] Furthermore, the digital twin virtual mapping module constructs a high-precision digital twin containing the system's geometric model, physical model, and behavioral model. Through this full-element digital twin, the module achieves real-time synchronous mapping between the physical oil circuit and the virtual model within the hydraulic system, making the internal operating status of the closed oil circuit visible. It can reproduce fault scenarios and simulate and verify maintenance plans, avoiding secondary damage to equipment caused by on-site trial and error maintenance. It can also perform virtual optimization of control strategies under different operating conditions, outputting optimal control parameters, solving the problems of fixed control parameters and poor adaptability to all operating conditions in traditional systems. At the same time, it provides simulation verification basis for system upgrades and transformations, improving the system's scalability and adaptability.

[0027] Furthermore, the hydraulic power unit includes an intelligent variable displacement piston pump with a built-in electronic control unit (ECU), an accumulator, and an oil temperature regulation unit. The ECU of the intelligent variable displacement plunger pump communicates bidirectionally with the collaborative control execution module, and can dynamically adjust the pump displacement according to instructions. The energy storage device is used for peak flow compensation and braking energy recovery in the system. The oil temperature regulating unit is connected to the central control unit and is used for closed-loop control of hydraulic oil temperature.

[0028] Furthermore, the electro-hydraulic control valve assembly includes an intelligent electronic proportional valve and a multi-position multi-way solenoid directional valve; The intelligent electronic proportional valve is equipped with a valve core displacement sensor and a drive control unit, which can receive PWM control commands to achieve continuous and precise regulation of flow and pressure. The multi-position multi-way electromagnetic directional valve is equipped with a valve position feedback sensor to provide real-time feedback on the valve core directional status in order to identify jamming faults.

[0029] Furthermore, the hydraulic actuator includes a chuck clamping hydraulic cylinder, a tool post indexing hydraulic cylinder, a tailstock sleeve hydraulic cylinder, and a spindle shifting hydraulic cylinder, all of which are equipped with built-in displacement and pressure sensors to collect piston rod displacement and working chamber pressure data in real time, thereby enabling monitoring of the operation status and fault identification of each actuator circuit.

[0030] A control method for an intelligent CNC hydraulic system, wherein the intelligent CNC hydraulic system comprises the following steps: S1: End-to-end data synchronous acquisition: Multi-dimensional time-domain operating parameters of key nodes in the hydraulic system are synchronously acquired through a multi-source heterogeneous sensor network and transmitted to the edge computing diagnostic node. S2: Edge data preprocessing and feature extraction. The edge computing diagnostic node performs filtering and noise reduction preprocessing on the raw data, extracts fault-sensitive feature indicators and completes local anomaly alarms, and then transmits the processed data to the dual-model fusion fault intelligent diagnosis module and the digital twin virtual mapping module respectively. S3: Dual-model fusion for intelligent fault diagnosis and tracing. It completes fault classification and location through the SVM sub-module, completes root cause tracing of faults by combining the FTA sub-module, and completes early fault warning and trend prediction through the LSTM sub-module. The results are synchronously transmitted to the corresponding modules. S4: Full lifecycle health management and operation and maintenance decision generation. Based on fault diagnosis results and component performance degradation models, it assesses the system health status, predicts the remaining service life, and generates corresponding hierarchical operation and maintenance strategies. S5: The digital twin virtual mapping module's digital twin synchronous mapping and simulation optimization uses real-time data to drive the synchronous mapping between the digital twin and the physical system, reproduce fault scenarios, and simulate and verify optimization solutions. The optimal solution is then output to the collaborative control execution module. S6: Adaptive Cooperative Control and Hierarchical Safety Protection. The cooperative control execution module dynamically adjusts the system control parameters based on the fault diagnosis results and the optimal control scheme, and executes hierarchical safety protection actions according to the severity level of the fault.

[0031] abnormal system pressure Fault phenomenon When a CNC lathe is in roughing operation, if the main oil circuit pressure does not reach the set value or the pressure fluctuation exceeds ±5%, abnormalities such as insufficient chuck clamping force and incomplete tool post locking will occur. These are the core typical faults preset in the document.

[0032] End-to-end detection process Pressure sensors installed at the pump inlet of the intelligent variable displacement plunger pump, the inlet and outlet of each valve in the electro-hydraulic control valve group, and each working oil chamber of the hydraulic actuator synchronously collect real-time pressure data of all oil circuit nodes; the main oil circuit flow sensor synchronously collects real-time flow data of the pipeline; the vibration sensor of the pump body shell collects the vibration signal of the pump body during operation; the temperature sensor and oil contamination sensor in the oil tank and the main return oil pipeline synchronously collect oil temperature and contamination data; All multi-dimensional time-domain operating parameters are transmitted to the edge computing diagnostic node in real time via the industrial bus; The edge computing diagnostic node uses a built-in adaptive variational Bayesian filtering algorithm to reduce noise in the original signal and suppress signal distortion caused by strong electromagnetic interference in the workshop. Then, through a multi-scale feature fusion algorithm, it extracts three types of fault-sensitive feature indicators from the pressure signal: fluctuation entropy value, frequency domain energy ratio from the flow signal, and wavelet packet energy entropy from the vibration signal. Simultaneously, it completes local primary threshold alarm and triggers a warning when the pressure exceeds the set range. Finally, the preprocessed feature data and the original time series data are transmitted to the SVM-LSTM dual-model fusion fault intelligent diagnosis module and the digital twin virtual mapping module, respectively.

[0033] Diagnostic and root cause investigation process The SVM submodule within the diagnostic module accurately classifies the fault type as "abnormal system pressure" based on the input feature indicators, and quickly locates the fault point range by combining the pressure and flow difference of each node; the built-in Fault Tree Analysis (FTA) submodule takes "abnormal system pressure" as the top event and traces back to the bottom event layer by layer, and completes root cause tracing by combining multi-source data, accurately distinguishing the fault into four types of causes: pump body wear, abnormal valve throttling, pipeline leakage, and abnormal oil viscosity. The LSTM submodule within the module predicts the development trend of pressure anomalies and anticipates the risk of fault deterioration based on long-term historical operating data, enabling early warning of faults 10-15 minutes in advance.

[0034] Control and Execution Process The full lifecycle health management module quantifies the health status of relevant components based on diagnostic results, classifies health levels (such as sub-health or minor faults in pump bodies), quantitatively predicts the remaining service life, and generates corresponding hierarchical operation and maintenance strategies for key monitoring or planned maintenance. The digital twin virtual mapping module drives a high-precision digital twin of all elements through real-time data, synchronously maps it with the physical system's operating conditions, reproduces the oil circuit pressure distribution and component operating status in a 1:1 ratio, simulates and reproduces the entire fault process, verifies the compensation effect of different control schemes, and outputs the optimal control parameters to the collaborative control execution module. The collaborative control execution module issues precise control commands to hardware components: If the volumetric efficiency of the intelligent variable displacement piston pump decreases, a command is sent to the pump's built-in ECU to dynamically increase the pump's displacement to compensate for pressure loss, while limiting the pump's maximum operating pressure to avoid overload and increased wear. If the valve port of the intelligent electronic proportional valve is worn, a PWM control command is sent to the valve's built-in drive control unit. Combined with the real-time feedback from the valve core displacement sensor inside the valve, the valve port opening is dynamically adjusted to compensate for throttling losses, thus forming a closed-loop control inside the valve. If the oil temperature is abnormal, a command is sent to the oil temperature regulation unit to start the heating or cooling function, stabilize the oil temperature within the preset range, and eliminate the influence of viscosity changes on pressure. If a serious fault is detected, such as leakage in the main pipeline or failure of the overflow valve, the graded safety protection will be immediately implemented. The multi-position multi-way solenoid directional valve will be controlled to cut off the oil supply to the corresponding circuit, and a command will be sent to the pump body ECU to reduce to an unloaded state, triggering an emergency stop warning to avoid safety accidents such as workpiece flying out or tool holder collision.

[0035] Multi-position multi-way solenoid directional valve stuck Fault phenomenon When a CNC lathe performs a tool change or spindle shift, if the tool post fails to reach its position or the spindle fails to shift between high and low speeds, the system will issue an action timeout alarm. This is one of the most frequent faults in the hydraulic system of a CNC lathe.

[0036] End-to-end detection process The valve position feedback sensor, which is matched with the multi-position multi-way solenoid directional valve, collects the valve core directional displacement signal in real time; the displacement sensor at the piston rod of the tool post indexing hydraulic cylinder or the spindle shifting hydraulic cylinder collects the hydraulic cylinder stroke data in real time; the pressure sensor at the inlet and outlet of the directional valve collects the pressure difference across the valve; and the vibration sensor on the valve body mounting surface collects the high-frequency vibration signal when the valve core moves. All data is transmitted to the edge computing diagnostic node in real time. After the node completes filtering and noise reduction, it extracts feature indicators such as valve core action delay time and vibration signal wavelet packet energy entropy. It identifies anomalies where the valve core action delay exceeds 0.3s, triggers a local primary alarm, and then transmits the data synchronously to the diagnostic module and the digital twin module.

[0037] Diagnostic and root cause investigation process The SVM submodule accurately classifies the fault as "reversing valve jamming" based on feature indicators and identifies the faulty valve component; the FTA analysis submodule combines oil contamination data and valve core action history data to trace the root cause to wear and jamming of the valve core mating surface caused by oil solid particle contamination. Based on the decay trend of valve core action characteristics, the LSTM submodule provides an early warning of jamming faults 12 minutes in advance and predicts the fault development path.

[0038] Control and Execution Process The full lifecycle health management module assesses the health level of the directional valve, generates corresponding operation and maintenance strategies, and simultaneously uploads fault information to the workshop MES system via the OPCUA protocol; The digital twin virtual mapping module reproduces the valve core jamming action process through a digital twin, simulates and verifies the effectiveness of emergency solutions such as reverse oil flow impact and pressure regulation, and outputs the optimal emergency control solution. The collaborative control execution module issues hierarchical control commands: If it is a slight early stage of jamming, a reverse reversing pulse command is sent to the reversing valve to control the valve core to briefly reverse and attempt to clear the jammed particles; at the same time, a command is sent to the pump body ECU to briefly increase the system oil supply pressure to assist the valve core in completing the reversing. Send commands to the oil temperature regulation unit to fine-tune the oil temperature, optimize the oil viscosity, and reduce the risk of valve core jamming; If there is severe jamming, immediately suspend the machining program, control the corresponding reversing valve to return to the neutral position to cut off the oil circuit, trigger an emergency stop warning, avoid equipment accidents such as tool holder collision and spindle gear breakage during shifting, and at the same time push standardized maintenance guidelines to the operation and maintenance terminal.

[0039] Wear of intelligent variable displacement plunger pump Fault phenomenon When the system is running unloaded, the pressure is normal, but after being loaded, the pressure drops rapidly and the flow is insufficient. During the processing, the actuator moves slowly and the output force is insufficient, and the pump body temperature continues to rise.

[0040] End-to-end detection process Pressure and flow sensors at the pump inlet and outlet collect real-time data on pump outlet pressure and output flow rate to calculate real-time volumetric efficiency; vibration and temperature sensors on the pump casing collect vibration signals and casing temperature during pump operation; and oil contamination sensors in the oil tank collect oil particle size data simultaneously. All data is transmitted to the edge computing diagnostic node in real time. After the node completes data preprocessing, it extracts indicators such as pump volumetric efficiency decay rate, pressure fluctuation entropy value, and high-frequency vibration impact characteristics. It identifies anomalies such as a pump volumetric efficiency decrease of 8%-15%, triggers a local early warning, and then transmits the data to subsequent modules.

[0041] Diagnostic and root cause investigation process The SVM submodule accurately classifies the fault as "variable pump wear," while the FTA analysis submodule traces the root cause to wear of the plunger distribution surface and bearing aging due to oil contamination. The LSTM submodule provides fault warnings 23 minutes in advance based on the pump performance degradation trend and quantitatively predicts the subsequent decline in volumetric efficiency. The diagnostic results are synchronously transmitted to the full life cycle health management module, which assesses the pump's health level as sub-healthy or with minor faults, predicts its remaining service life, and generates a planned operation and maintenance strategy.

[0042] Control and Execution Process The digital twin virtual mapping module simulates the effects of different displacement compensation strategies through a digital twin, verifies the parameter thresholds for safe operation of the pump, and outputs the optimal displacement compensation scheme. The collaborative control execution module sends adaptive adjustment commands to the pump's built-in ECU to dynamically adjust the pump's displacement according to the load conditions, and compensate for the flow and pressure losses caused by wear in real time. Under the premise of ensuring the system's output performance, it limits the pump's maximum speed and pressure to prevent further wear. Simultaneously, commands are sent to the oil temperature regulation unit to control the pump body return oil temperature within a reasonable range, reducing the risk of oil deterioration and accelerated component wear caused by high temperature; if the pump body wear is severe, a graded shutdown protection is immediately implemented to avoid systemic failure caused by pump body seizure.

[0043] Hydraulic cylinder internal leakage Fault phenomenon The workpiece becomes loose after being clamped by the chuck of the CNC lathe, and the tailstock sleeve slowly retracts after being tightened. The workpiece dimensions deviate beyond tolerance during the machining process. There is no external oil leakage. This is an internal fault that is extremely difficult to detect.

[0044] End-to-end detection process Pressure sensors in the rodless and rod chambers of the chuck clamping hydraulic cylinder or tailstock sleeve hydraulic cylinder collect real-time pressure change data between the two chambers; displacement sensors at the piston rod of the hydraulic cylinder collect real-time minute displacement signals of the piston rod; flow sensors at the inlet and outlet of the hydraulic cylinder collect the difference in inlet and outlet flow rates; and vibration sensors at the end of the cylinder body collect operating vibration signals. All data is transmitted to the edge computing diagnostic node in real time. The node extracts characteristic indicators such as the pressure difference between the two chambers, the flow loss rate, and the piston rod displacement drift, identifies abnormal leakage in the hydraulic cylinder, triggers a local early warning, and synchronously transmits the data to the diagnostic module.

[0045] Diagnostic and root cause investigation process The SVM submodule accurately locates the faulty hydraulic cylinder and classifies the fault as "internal leakage in the hydraulic cylinder"; the FTA analysis submodule traces the root cause to the seal failure caused by piston seal aging and wear of the inner wall of the cylinder. The full lifecycle health management module assesses the health level of the hydraulic cylinder, predicts the remaining service life of the seals, and generates corresponding operation and maintenance strategies.

[0046] Control and Execution Process The digital twin virtual mapping module uses a digital twin to simulate clamping force compensation schemes, verifying the clamping effect of different pressure compensation values ​​and avoiding the risk of workpieces flying out due to insufficient clamping force. The collaborative control execution module issues control commands and sends PWM signals to the intelligent electronic proportional valve drive control unit of the corresponding loop to dynamically adjust the valve opening, increase the oil supply pressure of the rodless chamber of the hydraulic cylinder, compensate for the pressure loss caused by internal leakage, and ensure that the chuck clamping force or tailstock clamping force meets the processing requirements. The maximum machining speed of the lathe is simultaneously limited, triggering a continuous warning to remind maintenance personnel to replace the seals in a planned manner; if the internal leakage is serious and the clamping force cannot be guaranteed, a safety shutdown is immediately executed, the oil supply to the chuck circuit is cut off, and the lathe is prohibited from starting the machining program.

[0047] The above-described embodiments are merely illustrative of certain implementations of the present invention, and are described in a relatively specific and detailed manner. However, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements are all within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. An intelligent numerical control hydraulic system, characterized in that: include The system comprises a hydraulic power unit, a hydraulic actuator unit, an electro-hydraulic control valve group, a sensor monitoring unit, and a central control unit. The hydraulic power unit is connected to the hydraulic actuator unit via the electro-hydraulic control valve group, the sensor monitoring unit is communicatively connected to the central control unit, and the central control unit is electrically connected to both the hydraulic power unit and the electro-hydraulic control valve group. The sensing and monitoring unit includes a multi-source heterogeneous sensor network, which is deployed at key nodes throughout the system to synchronously collect multi-dimensional operating parameters such as pressure, flow rate, temperature, vibration, displacement, and oil contamination. The central control unit integrates an edge computing diagnostic node, an SVM-LSTM dual-model fusion intelligent fault diagnosis module, a full lifecycle health management module, a digital twin virtual mapping module, and a collaborative control execution module; The edge computing diagnostic node is used for data preprocessing, feature extraction, and local anomaly alarm. The SVM-LSTM dual-model fusion fault intelligent diagnosis module includes an SVM sub-module and an LSTM sub-module, which are used for fault classification, localization and early warning. The full life cycle health management module is used for health status assessment, remaining life prediction, and generation of hierarchical operation and maintenance decisions. The digital twin virtual mapping module is used for fault scenario reproduction and virtual optimization of control strategies; The collaborative control execution module is used for adaptive adjustment of system control parameters and hierarchical safety protection.

2. The intelligent numerical control hydraulic system according to claim 1, characterized in that: The multi-source heterogeneous sensor network includes pressure sensors, flow sensors, temperature sensors, vibration sensors, displacement sensors, and oil contamination sensors; The pressure sensor and flow sensor are respectively installed at the pump port of the hydraulic power unit, the inlet and outlet of each valve of the electro-hydraulic control valve group, and each working oil chamber of the hydraulic actuator. The vibration sensors are installed in the pump body of the hydraulic power unit, the valve body of the electro-hydraulic control valve group, and the end of the cylinder body of the hydraulic actuator unit. The displacement sensor is installed at the valve core of the electro-hydraulic control valve group and the piston rod of the hydraulic actuator; The temperature sensor and the oil contamination sensor are installed in the system oil tank and the main return oil pipeline.

3. The intelligent numerical control hydraulic system according to claim 1, characterized in that: The edge computing diagnostic node incorporates an adaptive variational Bayesian filtering algorithm and a multi-scale feature fusion algorithm. The edge computing diagnostic node is also equipped with a local storage unit for caching system operation data and supporting encrypted transmission.

4. The intelligent CNC hydraulic system according to claim 1, characterized in that: The SVM-LSTM dual-model fusion fault intelligent diagnosis module also has a built-in fault tree FTA analysis submodule, which is preset with four types of typical faults: system vibration and abnormal noise, pressure abnormality, flow instability, and component failure.

5. The intelligent numerical control hydraulic system according to claim 1, characterized in that: The full lifecycle health management module divides the system health status into four levels: healthy, sub-healthy, minor fault, and serious fault.

6. The intelligent numerical control hydraulic system according to claim 1, characterized in that: The digital twin virtual mapping module constructs a high-precision digital twin with all elements, including the system's geometric model, physical model, and behavioral model.

7. The intelligent numerical control hydraulic system according to claim 1, characterized in that: The hydraulic power unit includes an intelligent variable displacement piston pump with a built-in electronic control unit (ECU), an accumulator, and an oil temperature regulation unit. The ECU of the intelligent variable displacement plunger pump communicates bidirectionally with the collaborative control execution module; The oil temperature regulating unit is connected to the central control unit.

8. The intelligent numerical control hydraulic system according to claim 1, characterized in that: The electro-hydraulic control valve group includes an intelligent electronic proportional valve and a multi-position multi-way solenoid directional valve. The intelligent electronic proportional valve is equipped with a valve core displacement sensor and a drive control unit. The multi-position multi-way electromagnetic directional valve is equipped with a valve position feedback sensor.

9. The intelligent numerical control hydraulic system according to claim 1, characterized in that: The hydraulic actuator includes a chuck clamping hydraulic cylinder, a tool post indexing hydraulic cylinder, a tailstock sleeve hydraulic cylinder, and a spindle shifting hydraulic cylinder, all compatible with the CNC lathe.

10. A control method for an intelligent numerical control hydraulic system, characterized in that, The intelligent numerical control hydraulic system based on any one of claims 1-9 includes the following steps: S1: End-to-end data synchronous acquisition: Multi-dimensional time-domain operating parameters of key nodes in the hydraulic system are synchronously acquired through a multi-source heterogeneous sensor network and transmitted to the edge computing diagnostic node. S2: Edge data preprocessing and feature extraction. The edge computing diagnostic node performs filtering and noise reduction preprocessing on the raw data, extracts fault-sensitive feature indicators and completes local anomaly alarms, and then transmits the processed data to the dual-model fusion fault intelligent diagnosis module and the digital twin virtual mapping module respectively. S3: Dual-model fusion for intelligent fault diagnosis and tracing. It completes fault classification and location through the SVM sub-module, completes root cause tracing of faults by combining the FTA sub-module, and completes early fault warning and trend prediction through the LSTM sub-module. The results are synchronously transmitted to the corresponding modules. S4: Full lifecycle health management and operation and maintenance decision generation. Based on fault diagnosis results and component performance degradation models, it assesses the system health status, predicts the remaining service life, and generates corresponding hierarchical operation and maintenance strategies. S5: The digital twin virtual mapping module's digital twin synchronous mapping and simulation optimization uses real-time data to drive the synchronous mapping between the digital twin and the physical system, reproduce fault scenarios, and simulate and verify optimization solutions. The optimal solution is then output to the collaborative control execution module. S6: Adaptive Cooperative Control and Hierarchical Safety Protection. The cooperative control execution module dynamically adjusts the system control parameters based on the fault diagnosis results and the optimal control scheme, and executes hierarchical safety protection actions according to the severity level of the fault.