Intelligent Adjustment System for Adaptive Multi-Environment Electric Control Valve Actuators of Ships

Abstract

The present invention discloses an intelligent adjustment system for an adaptive multi-environment marine electric control valve actuator, which relates to the technical field of control valve adjustment. The present invention includes an environment judgment module, a real-time monitoring module, an adaptive adjustment module, and a feedback optimization module; the environment judgment module collects environmental data through a variety of sensors distributed around the electric control valve and its periphery, and processes the data through signal conditioning and algorithms using a microprocessor. With the intelligent adjustment system for the adaptive multi-environment marine electric control valve actuator, the adaptability and stability of the marine electric control valve actuator in a changing environment are significantly improved; it can monitor and respond to environmental changes in real time, such as temperature, humidity, vibration, and electromagnetic interference, etc., ensuring that the electric control valve actuator can maintain precise control even under extreme working conditions, thereby improving the reliability and safety of the marine system.

Description

Technical Field

[0001] The present invention relates to the technical field of control valve regulation, and specifically to an intelligent regulation system for an adaptive multi-environment marine electric control valve actuator. Background Art

[0002] In the field of modern ship engineering, electric control valve actuators play a crucial role. They are metaphorically described as the "intelligent switches" of the ship's fluid transportation system, responsible for precisely regulating the flow rate and flow direction of various fluid media such as fuel oil, lubricating oil, seawater, and fresh water. These actuators are the core components to ensure the normal operation of many key systems on the ship, such as the power system, cooling system, ballast water system, etc. With the rapid development of ship technology towards large-scale, automated, and intelligent directions, the importance of electric control valve actuators in the complex and changing ship environment has become increasingly prominent;

[0003] Although electric control valve actuators play an important role in ship systems, in practical applications, they also expose many intractable problems. These problems include the contradiction between the demand for precise control and the actuator performance in the changing ship environment, as well as the challenge of maintaining stable operation under extreme working conditions. In addition, with the increase in the complexity of the ship environment, traditional electric control valve actuators have shown limitations in terms of adaptability, fault prediction, and self-regulation capabilities, and these problems directly affect the reliability and safety of ship systems.

[0004] To solve the above defects, the following technical solutions are provided. Summary of the Invention

[0005] The purpose of the present invention is to solve the problems of precise regulation and stability maintenance of electric control valve actuators in complex and changing environments in modern ships, and to propose an intelligent regulation system for an adaptive multi-environment marine electric control valve actuator.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] An intelligent regulation system for an adaptive multi-environment marine electric control valve actuator, comprising:

[0008] An environment judgment module, which collects environmental data through a variety of sensors distributed around the electric control valve, processes the data through signal conditioning and algorithms by a microprocessor, identifies the working conditions of the ship, and adapts to the working conditions of actuators at different positions through differential monitoring, and optimizes the sensor layout;

[0009] The real-time monitoring module tracks the operating parameters of the electric control valve, and obtains and transmits the motor current, valve opening displacement, and motor speed data in real time; at the same time, it interacts with the environment judgment module to construct a multi-variable correlation monitoring matrix, analyzes the data according to predefined rules and data mining and deep learning technologies, collaboratively judges the fault risk, and provides comprehensive data support for the adjustment decision-making;

[0010] The adaptive adjustment module matches the initial strategy from the strategy library according to the working conditions identified by the environment judgment module. The strategy library stores multiple control parameter sets classified by working conditions and continuously optimizes them. It evaluates the performance by collecting operation data, analyzes the deviation to adjust the strategy, predicts the environmental changes at the same time, verifies the pre-adjusted strategy, optimizes the strategy using machine learning and reinforcement learning, and controls the actuator motor through the power drive unit to achieve valve adaptation adjustment;

[0011] The feedback optimization module compares the actual operating parameters of the actuator with the adjustment instruction target in real time, corrects the deviation with the help of a PID controller, and realizes closed-loop control to ensure accurate operation; accumulates operation data periodically, trains and optimizes using the LSTM model, mines the adjustment rules of working conditions, and iteratively optimizes the strategy library and the parameters of the adaptive algorithm in reverse to improve the adaptability of the system to the complex environment of the ship.

[0012] Furthermore, the execution process of the environment judgment module is as follows:

[0013] It is distributed throughout the electric control valve and surrounding points to construct a multi-dimensional perception system, including: the temperature sensor uses a platinum resistance temperature detector; the vibration sensor selects a micro-electromechanical system accelerometer; the electromagnetic interference monitoring uses a broadband loop antenna with a spectrum analyzer;

[0014] After the data collected by each sensor is amplified, filtered, and converted from analog to digital and summarized through the signal conditioning circuit, it is input into the microprocessor with built-in algorithms;

[0015] Using big data clustering analysis and the convolutional neural network algorithm in deep learning, analyze the short-term, medium-term, and long-term environmental data, train the convolutional neural network model with typical ship working condition samples, identify the current working condition classification, and anchor the basic situation for subsequent adjustment;

[0016] Implement differential monitoring paths to adapt to the complex ship structure; optimize the sensor layout according to different location characteristics.

[0017] Furthermore, the implementation of the differential monitoring path by the environment judgment module is as follows:

[0018] Each electric control valve is given a unique identity code, and the coding rules are integrated into the position information. The sensor data is automatically classified and stored according to the code, and an independent data set is constructed, so that the data of the actuators at each position are in their place; when classifying electric control valves at different positions, they are distinguished by analyzing the similarity of the electric control valve monitoring data. The specific process is as follows:

[0019] Clean the monitoring data collected from each electric control valve actuator; use statistical methods to identify and eliminate outliers; at the same time, check the integrity of the data, and use linear interpolation or mean filling methods to supplement the missing data points; and normalize the dimensions and value ranges of different parameters through the minimum-maximum normalization method;

[0020] For the monitoring data vector of the electric control valve, let the monitoring data vector of the electric control valve A be , the monitoring data vector of electric control valve B is , where T, H, V, and A are the four parameters of temperature, humidity, vibration, and electromagnetic interference respectively;

[0021] The similarity of the monitoring data of electric control valve A and electric control valve B is calculated by the formula: , where is the cosine similarity of the monitoring data of electric control valve A and electric control valve B, and the value range is ;

[0022] Each electric control valve is regarded as a separate class. By calculating the cosine similarity between each pair of electric control valves, a pair of electric control valves with the highest cosine similarity is selected and merged into a new class. This process is repeated, and classes that meet the similarity standard are continuously merged until the preset stop condition is reached. The preset stop condition includes that the number of classes reaches a preset value or the similarity between classes is lower than a preset threshold, so as to obtain a hierarchical classification result, from the finest granularity of each electric control valve as a class to the final merging into several large categories; a unified control strategy is implemented for electric control valves of different categories.

[0023] Furthermore, the specific operation steps of the real-time monitoring module are as follows:

[0024] Electric control valve operation parameter tracking: Focus on the operation indicators of the electric control valve, by embedding current sensors in the motor windings, installing displacement encoders on the transmission chain or ball screw, and configuring speed sensors on the motor shaft ends, to obtain motor current, valve opening displacement, and motor speed parameters in real time. The data is sampled and updated at a preset frequency every second, and directly transmitted to the adaptive adjustment module in the form of digital signals to control the instant working status of the electric control valve;

[0025] Synchronous monitoring of environmental auxiliary variables: interact with the environmental judgment module data, continuously pay attention to the dynamics of temperature, humidity, vibration, and electromagnetic interference environmental parameters, combine the actuator operating parameters, build a multi-variable correlation monitoring matrix, and provide comprehensive and three-dimensional data support for adjustment decisions.

[0026] Furthermore, the specific operation steps of synchronous monitoring of environmental auxiliary variables in the real-time monitoring module are as follows:

[0027] After receiving the temperature, humidity, vibration amplitude and frequency, electromagnetic interference spectrum and intensity data from the environment judgment module, as well as the electric control valve motor current, valve opening displacement, and motor speed parameters collected by itself, various types of data are sorted in a unified format; each data is given a timestamp, data source identifier, and key meta-information of physical quantity units;

[0028] Construct a two-dimensional matrix structure, where the row dimension corresponds to the time series and the column dimension divides different monitoring variables, listing the temperature, humidity, vibration axial data, electromagnetic interference index, motor current, valve opening, and motor speed in turn;

[0029] The real-time collected and formatted data are filled into the matrix according to the corresponding row and column positions, forming a data puzzle that is dynamically updated over time and integrates multiple variables, presenting the interactive changes of various parameters at different times;

[0030] Built-in predefined fault association rules. When the monitoring matrix data is updated, the rule conditions are compared in sequence. Once a match is made, a primary fault warning signal is immediately generated, with relevant parameter data details for subsequent in-depth analysis.

[0031] Data mining algorithms and deep learning models are used to analyze historical data of the monitoring matrix that triggers early warnings or is accumulated over a long period of time. The nonlinear relationship between different environmental factors and actuator operation failures is explored to verify and refine the fault risk category and severity assessment, and a comprehensive diagnostic report containing fault type, location, and estimated fault time range is output, providing a detailed basis for the adaptive adjustment module to implement precise policies.

[0032] Furthermore, the specific operation steps of the adaptive adjustment module are as follows:

[0033] Strategy library pre-storage and intelligent matching: Relying on the storage unit, the electric control valve regulation strategy library under the full spectrum of ship working conditions is built-in, and the motor drive voltage and current curves, valve opening adjustment step length, frequency, and speed limit parameter sets are stored according to the working conditions; the regulation strategy library is continuously optimized;

[0034] Identify the working conditions based on the environmental judgment module and match the initial strategy blueprint;

[0035] Dynamic Fine-tuning Mechanism: Integrating fuzzy logic and particle swarm optimization algorithm, taking the data of the real-time monitoring module as dynamic input. Fuzzy logic quantifies the environmental and operating fuzzy quantities into control weights, and the particle swarm optimization algorithm real-time fine-tunes the strategy parameters driven by the optimization objective, collaboratively generating the optimal adjustment instructions, which control the operation of the actuator motor through the power drive unit to achieve flexible and adaptive adjustment of the valve.

[0036] Furthermore, the specific operation steps for continuously optimizing the adjustment strategy library in the adaptive adjustment module are as follows:

[0037] During the operation of the ship, continuously collect the actual operation data of the electric control valve actuator, including environmental parameters, actuator operation parameters, and data related to the adjustment effect; preprocess the collected data to remove outliers and noise interference; at the same time, normalize the data so that the data of different parameters are on the same magnitude;

[0038] Regularly evaluate the performance of the electric control valve actuator, and calculate the deviation between the actual operation data and the ideal performance index according to the preset performance index;

[0039] Analyze the reasons for the deviation to determine whether it is caused by environmental changes, actuator aging, or unreasonable control strategies; through statistical analysis of the operation data, find out the key factors related to the deviation;

[0040] According to the results of performance evaluation and deviation analysis, make targeted adjustments to the parameters in the adjustment strategy library; for the motor drive parameters, optimize according to the actual operation state of the motor;

[0041] Adopt time series analysis method to monitor and predict the change trends of environmental parameters such as temperature, humidity, vibration, and electromagnetic interference, and combine with factors such as the ship's navigation route and seasonal changes to anticipate the extreme environmental conditions that may occur in advance;

[0042] According to the environmental trend prediction results, pre-adjust the relevant strategies in the adjustment strategy library before the actual environmental change; after the strategy pre-adjustment, verify the new strategy using the simulation environment. By establishing a mathematical model of the electric control valve, simulate the operation of the electric control valve under different environmental conditions, and evaluate the effectiveness and feasibility of the pre-adjusted strategy;

[0043] When the actual environment changes and the pre-adjusted strategy is verified, real-time online update the strategy parameters in the adjustment strategy library, and ensure that the actuator control system can stably switch to the new control strategy; during the switching process, monitor the operation state of the actuator to prevent instability caused by strategy switching;

[0044] Collect the operation data of the electric control valve actuator under different working conditions, including normal operation data and fault status data, as the training samples of the machine learning model; divide these data into training set, validation set and test set, select machine learning algorithms, including neural network, support vector machine or decision tree, and construct a performance prediction model for the electric control valve actuator;

[0045] Use the environmental parameters and actuator operation parameters as input features, and the valve control effect index as the output target to train the model; through the trained machine learning model, predict and optimize the optimal control strategy under different working conditions; the model outputs the recommended motor drive voltage, current curve, valve opening adjustment step, frequency, and speed limit parameters according to the real-time input environmental and operation data;

[0046] Utilize the reinforcement learning algorithm to enable the electric control valve actuator to explore and learn the optimal control strategy during actual operation; by setting a reward function, encourage the actuator to optimize the performance index on the premise of meeting the control requirements;

[0047] As the ship operation time increases and new data accumulates, regularly update the machine learning model to adapt to the changing working conditions and actuator performance; adopt the model fusion technology to fuse the prediction results of multiple different machine learning models to obtain better control strategy suggestions.

[0048] Further, the specific operation steps for evaluating the effectiveness and feasibility of the pre-adjustment strategy in the adaptive adjustment module are as follows:

[0049] Conduct a comprehensive analysis through the evaluation parameters of the pre-adjustment strategy. The evaluation parameters include:

[0050] Valve opening error: Calculate the average value gt and standard deviation sd of the difference between the actual opening and the target opening of the electric control valve under simulated different environmental working conditions;

[0051] Flow control error: By establishing a mathematical relationship model between flow and valve opening, calculate the theoretical flow according to the simulated valve opening, and compare it with the actual simulated flow value to obtain the flow control error wf;

[0052] Response time: Record the time required for the electric control valve to reach 90% of the target opening from receiving the control signal; when simulating different environmental working conditions, compare the changes in the response time and calculate the average response time he;

[0053] Overshoot: Calculate the ratio kb of the maximum deviation value exceeding the target opening to the target opening during the adjustment process of the electric control valve, expressed as a percentage;

[0054] Substitute the obtained average opening difference value gt, standard deviation of opening difference sd, average response time he, and opening ratio kb after normalization into the following formula: To obtain the comprehensive judgment value FD, where are the preset weight coefficients of the average opening difference value gt, standard deviation of opening difference sd, average response time he, and opening ratio kb respectively, and use the obtained comprehensive judgment value FD as the standard to measure the effectiveness and feasibility of the adjustment strategy; when the comprehensive judgment value FD does not exceed the preset comprehensive judgment threshold, it is determined that the effectiveness and feasibility of the pre-adjustment strategy meet the standards.

[0055] Furthermore, the execution process of the feedback optimization module is as follows:

[0056] Deviation real-time correction: Compare the actual operating parameters of the actuator with the preset target of the adjustment instruction. With the help of a proportional-integral-derivative controller, once the deviation exceeds the allowable threshold, calculate the correction amount immediately according to the proportional-integral-derivative algorithm, correct the adjustment instruction and send it back to the adaptive adjustment module, and the closed-loop control ensures that the operation accurately tracks the target;

[0057] Self-learning evolution: Accumulate operation data according to voyages or monthly cycles, feed it into the long short-term memory network model of machine learning for training and optimization, mine the working condition - adjustment rules, and iteratively optimize the strategy library and adaptive algorithm parameters in reverse to improve the system's adaptability to the complex environment of the ship, extend the equipment life, and increase the efficiency of ship operation and maintenance.

[0058] Compared with the prior art, the beneficial effects of the present invention are:

[0059] In the present invention, through the intelligent adjustment system of the adaptive multi-environment ship electric control valve actuator, the adaptability and stability of the ship electric control valve actuator in a changing environment are significantly improved; it can monitor and respond to environmental changes in real time, such as temperature, humidity, vibration, and electromagnetic interference, etc., to ensure that the electric control valve actuator can maintain precise control under extreme working conditions, thereby improving the reliability and safety of the ship system;

[0060] In the present invention, using machine learning and reinforcement learning algorithms, deeply analyze and learn the operation data of the electric control valve actuator, optimize the control strategy, and predict potential failure risks; this self-learning and adaptive ability enables the system to continuously evolve, improve performance, while reducing the occurrence of failures, extending the equipment life, and enhancing the efficiency of ship operation and maintenance;

[0061] In the present invention, through closed-loop control and real-time feedback optimization, it ensures the precise operation of the electric control valve actuator, reduces the adjustment error and response time; periodically accumulates operation data, iteratively optimizes the strategy library and adaptive algorithm parameters in reverse, further improves the system's adaptability to complex environments, realizes the increase of ship operation and maintenance efficiency, reduces the maintenance cost, and improves the economic efficiency of ship operation. Brief Description of the Drawings

[0062] For the convenience of those skilled in the art to understand, the present invention will be further described below in conjunction with the drawings;

[0063] Figure 1 It is the overall system block diagram of the present invention. Detailed Embodiments

[0064] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0065] It should be understood that the terms "including" and "comprising" used in the specification and claims of this disclosure indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0066] It should also be understood that the terms used in the specification of this disclosure are only for the purpose of describing specific embodiments and are not intended to limit this disclosure. As used in the specification and claims of this disclosure, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms. It should also be further understood that the term "and / or" used in the specification and claims of this disclosure refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0067] As Figure 1 shown, the intelligent adjustment system of the adaptive multi-environment ship electric control valve actuator includes an environment judgment module, a real-time monitoring module, an adaptive adjustment module, and a feedback optimization module;

[0068] The environment judgment module collects environmental data through a variety of sensors distributed around the electric control valve and its periphery, processes the data through signal conditioning and algorithms using a microprocessor, identifies the working conditions of the ship, provides a basic scenario for subsequent adjustments, adapts to the working conditions of actuators at different positions through differential monitoring, and optimizes the sensor layout at the same time;

[0069] Distributed around the electric control valve and its peripheral points, a multi-dimensional perception system is constructed, including:

[0070] The temperature sensor uses a platinum resistance temperature detector, which has excellent linearity, strong stability, and a wide range, and can accurately capture changes in the temperature of the environment. The humidity sensor uses a capacitive humidity-sensitive element, which responds sensitively to the relative humidity range of 20%-95%, and provides real-time feedback on the fluctuations in water vapor content. The vibration sensor uses a MEMS (micro-electromechanical system) accelerometer, which can sense vibrations in the three axes of X, Y, and Z, with a range of 0-10g (g is the acceleration of gravity), which is suitable for ship shaking and equipment vibration conditions. Electromagnetic interference monitoring uses a wide-band loop antenna with a high-precision spectrum analyzer to effectively lock the interference intensity and spectrum characteristics of the 10kHz-2GHz electromagnetic frequency band, and gain a full understanding of the electromagnetic environment. The data collected by each sensor is summarized by a signal conditioning circuit (to achieve amplification, filtering, and analog-to-digital conversion) and then input into a microprocessor with a built-in algorithm.

[0071] Using big data cluster analysis and convolutional neural network (CNN) algorithms in deep learning, we deeply analyze short-term (several minutes), medium-term (several hours), and long-term (several days) environmental data, train CNN models with typical ship operating condition samples, and identify current operating conditions, including "high temperature, high humidity, strong vibration, high electromagnetic interference" and "low temperature, low humidity, stable and low interference", etc., to lay the foundation for subsequent adjustments.

[0072] Implement differentiated monitoring implementation paths to adapt to complex ship structures. Electric control valves are widely distributed, covering different areas such as engine rooms, decks, and cabins, and the working conditions at each location are very different; specifically:

[0073] Each electric control valve is given a unique identity code, and the coding rules are integrated into the position information. The sensor data is automatically classified and stored according to the code to build an independent data set, so that the actuator data at each position is in its place to avoid confusion; when classifying electric control valves at different positions, they are distinguished by analyzing the similarity of the electric control valve monitoring data. The specific process is as follows:

[0074] Clean the monitoring data collected from each electric control valve actuator. Due to the complex ship environment, the sensor may be disturbed and produce abnormal values; use statistical methods to identify and eliminate abnormal values; at the same time, check the integrity of the data, and use linear interpolation or mean filling to supplement the missing data points; and normalize the dimensions and value ranges of different parameters through the minimum-maximum normalization method; for the monitoring data vector of the electric control valve, let the monitoring data vector of the electric control valve A be , the monitoring data vector of electric control valve B is , where T, H, V, and A are temperature, humidity, vibration, and electromagnetic interference, respectively; the similarity of the monitoring data of electric control valve A and electric control valve B is calculated by the formula: , where is the cosine similarity of the monitoring data of electric control valve A and electric control valve B, and its value range is , the closer the value is to 1, it indicates that the monitoring data of the two electric control valves are more similar in direction, that is, the working conditions are more similar; regarding each electric control valve as a separate class, select a pair of electric control valves with the highest cosine similarity by calculating the cosine similarity between each pair of electric control valves, and merge them into a new class. Repeat this process and continuously merge the classes whose similarity reaches the standard until the preset stop condition is reached, including the number of classes reaching the preset value or the similarity between classes being lower than the preset threshold, so as to obtain a hierarchical classification result, from each electric control valve at the finest granularity as a class to finally merging into several large classes; implement a unified control strategy for electric control valves of different classes;

[0075] Optimize the sensor layout according to different location characteristics; densely install temperature and vibration sensors around the cabin actuator, close to the heat source and the equipment vibration source; for the deck actuator, while emphasizing the waterproof and wind-sand protection design, improve the protection level of vibration and temperature sensors to cope with harsh climates; for the cabin actuator, focus on high-precision temperature and humidity sensors to meet the requirements of a comfortable environment, and the electromagnetic interference monitoring is adapted to the low-noise frequency band to fit the indoor electrical environment.

[0076] The real-time monitoring module tracks the operating parameters of the electric control valve, and obtains and transmits the motor current, valve opening displacement, and motor speed data in real time; at the same time, it interacts with the environment judgment module to construct a multi-variable correlation monitoring matrix, analyzes the data according to predefined rules and data mining and deep learning technologies, collaboratively judges the fault risk, and provides comprehensive data support for the adjustment decision-making;

[0077] Tracking of electric control valve operating parameters: Focus on the core operating indicators of the electric control valve. By embedding a current sensor in the motor winding (based on the Hall effect principle, with an accuracy of milliamperes), installing a high-precision displacement encoder on the transmission chain or ball screw (with a resolution of 0.01 mm), and configuring a speed sensor at the motor shaft end (using a magnetoelectric type, with a speed measurement accuracy of 0.1%), the motor current, valve opening displacement, and motor speed parameters are obtained in real time, and the data is sampled and updated at a preset frequency per second, and directly transmitted to the adaptive adjustment module in digital signal form to control the immediate working state of the electric control valve;

[0078] Synchronous monitoring of environmental auxiliary variables: Interact with the environment judgment module data, continuously monitor the dynamics of environmental parameters such as temperature, humidity, vibration, and electromagnetic interference, and combine the operating parameters of the actuator to construct a multi-variable correlation monitoring matrix to provide comprehensive and three-dimensional data support for the adjustment decision-making. For example, if a sudden temperature rise is detected along with a valve action lag, analyze the potential associated fault risks collaboratively; the specific process is as follows:

[0079] After receiving data such as temperature, humidity, vibration amplitude and frequency, electromagnetic interference spectrum and intensity from the environmental judgment module, as well as the motor current of the electric control valve, valve opening displacement, and motor speed parameters collected by itself, it organizes various types of data in a unified format. Assigns key meta-information such as time stamps, data source identifiers (to distinguish whether they are environmental parameters or actuator operation parameters), and physical quantity units to each data for subsequent efficient processing and interpretation; constructs a two-dimensional matrix structure, where the row dimension corresponds to the time series (arranging data row by row according to the sampling order, and each row represents all parameters collected at the same moment), and the column dimension divides different monitoring variables, listing temperature, humidity, vibration data of each axis, key indicators of electromagnetic interference, motor current, valve opening, motor speed, etc. in sequence. Fills the real-time collected and formatted data into the matrix according to the corresponding row and column positions, forming a "data puzzle" that dynamically updates over time and integrates multiple variables, intuitively presenting the interactive change trend of each parameter at different times;

[0080] Built-in a series of predefined fault correlation rules, such as "When the environmental temperature rises by more than 20°C within 10 minutes and the valve opening action speed slows down by more than 50% compared to the normal average value, mark it as a potential risk of thermal impact failure" and "When the electromagnetic interference intensity exceeds the threshold by 5 dB in a specific frequency band and the motor current fluctuation amplitude exceeds 10% of the rated value, prompt the risk of electromagnetic compatibility abnormality" and other rule sets. When monitoring the update of the matrix data, the system sequentially compares each rule condition. Once a match is found, it immediately generates a primary fault warning signal, accompanied by the details of relevant suspicious parameter data for subsequent in-depth analysis; for the historical data of the monitoring matrix that triggers warnings or accumulates over a long time, uses data mining algorithms (the Apriori algorithm for mining association rules to mine frequently co-occurring abnormal parameter combinations) and deep learning models (constructs a fault prediction model based on the long short-term memory network LSTM to learn the hidden fault patterns in time series data) to conduct in-depth analysis; mines the potential non-linear relationship between different environmental factors and actuator operation failures, verifies and refines the fault risk categories and severity assessments, and outputs a comprehensive diagnostic report including the fault type (mechanical jamming, electrical performance degradation, environmental adaptability problems, etc.), possible occurrence locations (motor, transmission components, valve seals, etc.), and the estimated fault time range, providing detailed basis for the adaptive adjustment module to take precise measures.

[0081] The adaptive adjustment module matches the initial strategy from the strategy library according to the working conditions identified by the environmental judgment module. The strategy library stores multiple control parameter sets classified by working conditions and continuously optimizes them. By collecting operation data to evaluate performance, analyzing deviations to adjust strategies, predicting environmental changes, and verifying pre-adjusted strategies, and using machine learning and reinforcement learning to optimize strategies, it controls the actuator motor through the power drive unit to achieve valve adaptation adjustment;

[0082] Policy Library Pre-storage and Intelligent Matching: Relying on the storage unit, an electric control valve adjustment policy library under the full operating conditions spectrum of the ship is built-in, and the motor drive voltage and current curves, valve opening adjustment step sizes, frequencies, and speed limit parameter sets are stored in a refined manner according to the operating conditions; continuously optimize the adjustment policy library, and the specific process is as follows:

[0083] During the operation of the ship, continuously collect the actual operation data of the electric control valve actuator, including environmental parameters (such as temperature, humidity, vibration, electromagnetic interference, etc.), actuator operation parameters (such as motor current, valve opening, motor speed, etc.), and data related to the adjustment effect (such as valve control accuracy, response time, stability, etc.); preprocess the collected data to remove outliers and noise interference; at the same time, normalize the data to make the data of different parameters on the same order of magnitude for subsequent analysis and comparison; regularly (such as every hour or every shift) evaluate the performance of the electric control valve actuator, and calculate the deviation between the actual operation data and the ideal performance index according to the preset performance indicators (such as the valve opening error within ±1%, the response time less than 5 seconds, etc.); analyze the reasons for the deviation to determine whether it is caused by environmental changes, actuator aging, unreasonable control strategies or other factors. Through the statistical analysis of a large amount of operation data, find out the key factors related to the deviation;

[0084] According to the results of the performance evaluation and deviation analysis, make targeted adjustments to the parameters in the adjustment policy library; if it is found that the valve control accuracy decreases under a certain operating condition, adjust the valve opening adjustment step size and frequency parameters; for the motor drive parameters, such as voltage and current curves, optimize them according to the actual operating state of the motor; use time series analysis methods (such as ARIMA model) to monitor and predict the change trends of environmental parameters such as temperature, humidity, vibration, and electromagnetic interference, and combine factors such as the ship's navigation route and seasonal changes to predict the extreme environmental operating conditions in advance; according to the environmental trend prediction results, pre-adjust the relevant strategies in the adjustment policy library before the actual environmental changes; after the strategy pre-adjustment, verify the new strategy using the simulation environment; by establishing a mathematical model of the electric control valve, simulate the operation of the electric control valve under different environmental operating conditions, and evaluate the effectiveness and feasibility of the pre-adjusted strategy; specifically, conduct a comprehensive analysis through the evaluation parameters of the pre-adjustment strategy, and the evaluation parameters include:

[0085] Valve opening error: Calculate the average value gt and standard deviation sd of the difference between the actual opening and the target opening of the electric control valve under different simulated environmental conditions; Flow control error: By establishing a mathematical relationship model between flow and valve opening (based on fluid mechanics principles and valve characteristic curves), calculate the theoretical flow according to the simulated valve opening, and compare it with the actual simulated flow value to obtain the flow control error wf; Response time: Record the time required for the electric control valve to reach 90% (or other set ratio) of the target opening from the receipt of the control signal; When simulating different environmental conditions, compare the changes in the response time and calculate the average response time he; Overshoot: Calculate the ratio kb of the maximum deviation value exceeding the target opening to the target opening during the adjustment process of the electric control valve, expressed as a percentage; Normalize the obtained average opening difference value gt, opening difference standard deviation sd, average response time he, and opening ratio kb and substitute them into the following formula: To obtain the comprehensive judgment value FD, where Are the preset weight coefficients of the average opening difference value gt, opening difference standard deviation sd, average response time he, and opening ratio kb respectively, and use the obtained comprehensive judgment value FD as the standard to measure the effectiveness and feasibility of the adjustment strategy; When the comprehensive judgment value FD does not exceed the preset comprehensive judgment threshold, it is judged that the effectiveness and feasibility of the pre-adjustment strategy meet the standards.

[0086] When the actual environment changes and the pre-adjustment strategy is verified, the strategy parameters in the adjustment strategy library are updated online in real time, and it is ensured that the actuator control system can stably switch to the new control strategy; During the switching process, monitor the operating state of the actuator to prevent short-term instability caused by strategy switching;

[0087] Collect a large amount of operation data of the electric control valve actuator under different working conditions, including normal operation data and fault state data, as the training samples of the machine learning model. Divide these data into a training set, a validation set and a test set, with the ratio set as 7:2:1; select a suitable machine learning algorithm, such as neural network (such as multi-layer perceptron MLP), support vector machine (SVM) or decision tree, etc., to build a performance prediction model of the electric control valve actuator. Use environmental parameters and actuator operation parameters as input features, and use valve control effect indicators (such as control accuracy, response time, etc.) as output targets to train the model; through the trained machine learning model, predict and optimize the optimal control strategy under different working conditions. The model can output parameters such as the recommended motor drive voltage and current curves, valve opening adjustment step size, frequency, and speed limit according to the real-time input environmental and operation data; use reinforcement learning algorithms (such as Q-learning or deep Q-network DQN) to enable the electric control valve actuator to continuously explore and learn the optimal control strategy during actual operation. By setting a reward function, encourage the actuator to optimize performance indicators (such as reducing energy consumption, improving stability, etc.) on the premise of meeting the control requirements; as the ship operation time increases and new data accumulates, regularly update the machine learning model to adapt to the changing working conditions and actuator performance; adopt model fusion technology to fuse the prediction results of multiple different machine learning models (such as neural network models and decision tree models), and comprehensively consider the advantages of each model to obtain more reliable control strategy suggestions.

[0088] The environment judgment module identifies the working conditions and matches the initial strategy blueprint, including moderately reducing the motor drive voltage to prevent overheating under high-temperature working conditions, and narrowing the valve adjustment step size to ensure stability in a high-vibration environment;

[0089] Dynamic fine-tuning mechanism: Integrate fuzzy logic and particle swarm optimization algorithm (PSO), use the data of the real-time monitoring module as dynamic input, fuzzy logic quantifies the environmental and operation "fuzzy quantities" (such as temperature "hot or cold", vibration "strong or weak", etc.) into control weights, and the PSO algorithm drives the real-time fine-tuning of strategy parameters with the optimization target, and collaboratively generates the optimal adjustment instruction, and controls the operation of the actuator motor through the power drive unit (builds an efficient inverter circuit with insulated gate bipolar transistors IGBT to adapt to the ship power supply), achieving flexible and adaptive adjustment of the valve.

[0090] The feedback optimization module compares the actual operation parameters of the actuator with the adjustment instruction target in real time, corrects the deviation with the help of a PID controller, and realizes closed-loop control to ensure accurate operation; accumulates operation data according to the cycle, uses the LSTM model for training and optimization, excavates the working condition adjustment rules, and iteratively optimizes the strategy library and adaptive algorithm parameters in reverse to improve the system's adaptability to the complex ship environment;

[0091] Real-time deviation correction: Compare the actual operating parameters of the actuator (valve opening, movement speed, etc.) with the preset target of the adjustment instruction. With the help of the proportional-integral-differential (PID) controller, once the deviation exceeds the allowable threshold, the deviation correction amount is calculated in real time according to the PID algorithm, and the adjustment instruction is corrected and sent back to the adaptive adjustment module. Closed-loop control ensures accurate tracking of the operation target and copes with real-time fluctuations in complex working conditions.

[0092] Long-term self-learning evolution: accumulate operating data (adjustment errors of various working conditions, response time, fault frequency, etc.) by voyage or monthly cycle, feed the long short-term memory network (LSTM) model training optimization of machine learning, explore the deep laws of working condition-adjustment, reverse iterate and optimize the strategy library and adaptive algorithm parameters, and improve the system's "response intelligence" to the complex environment of ships year by year, extend equipment life, and improve the efficiency of ship operation and maintenance.

[0093] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. Intelligent adjustment system for adaptive multi-environment ship electric control valve actuator, characterized in that: include: The environmental judgment module collects environmental data through various sensors distributed around the electric control valve, and identifies the working conditions of the ship through signal conditioning and microprocessor algorithm processing. It also optimizes the sensor layout by implementing differentiated monitoring to adapt to the working conditions of actuators in different positions; The real-time monitoring module tracks the operating parameters of the electric control valve, acquires and transmits the motor current, valve opening displacement and motor speed data in real time; it interacts with the environmental judgment module to build a multivariable correlation monitoring matrix, analyzes data based on predefined rules and data mining and deep learning technologies, jointly judges fault risks, and provides comprehensive data support for adjustment decisions; The specific operation steps of the real-time monitoring module are as follows: Electric control valve operation parameter tracking: Focus on the operation indicators of the electric control valve, by embedding current sensors in the motor windings, installing displacement encoders on the transmission chain or ball screw, and configuring speed sensors on the motor shaft ends, to obtain motor current, valve opening displacement, and motor speed parameters in real time. The data is sampled and updated at a preset frequency every second, and directly transmitted to the adaptive adjustment module in the form of digital signals to control the instant working status of the electric control valve; Synchronous monitoring of environmental auxiliary variables: interact with the environmental judgment module data, continuously pay attention to the dynamics of temperature, humidity, vibration, and electromagnetic interference environmental parameters, combine the actuator operation parameters, build a multi-variable correlation monitoring matrix, and provide comprehensive and three-dimensional data support for adjustment decisions; The specific operation steps of synchronous monitoring of environmental auxiliary variables in the real-time monitoring module are as follows: After receiving the temperature, humidity, vibration amplitude and frequency, electromagnetic interference spectrum and intensity data from the environment judgment module, as well as the electric control valve motor current, valve opening displacement, and motor speed parameters collected by itself, various types of data are sorted in a unified format; each data is given a timestamp, data source identifier, and key meta-information of physical quantity units; Construct a two-dimensional matrix structure, with the row dimension corresponding to the time series and the column dimension dividing different monitoring variables, listing the temperature, humidity, vibration axial data, electromagnetic interference index, motor current, valve opening, and motor speed in turn; The real-time collected and formatted data are filled into the matrix according to the corresponding row and column positions, forming a data puzzle that is dynamically updated over time and integrates multiple variables, presenting the interactive changes of various parameters at different times; Built-in predefined fault association rules. When the monitoring matrix data is updated, the rule conditions are compared in sequence. Once a match is made, a primary fault warning signal is immediately generated, with relevant parameter data details for subsequent in-depth analysis. Use data mining algorithms and deep learning models to analyze historical data of monitoring matrices that trigger early warnings or accumulate over a long period of time; mine the nonlinear relationship between different environmental factors and actuator operation failures, verify and refine the fault risk category and severity assessment, and output a comprehensive diagnostic report that includes the fault type, location, and estimated fault time range; The adaptive adjustment module matches the initial strategy from the strategy library according to the working conditions identified by the environmental judgment module. The strategy library stores a variety of control parameter sets by working conditions and continuously optimizes them. It evaluates performance and analyzes deviation adjustment strategies by collecting operating data. It also predicts environmental changes and verifies pre-adjustment strategies. It uses machine learning and reinforcement learning to optimize strategies and controls the actuator motor through the power drive unit to achieve valve adaptive adjustment. The feedback optimization module compares the actual operating parameters of the actuator with the adjustment instruction targets in real time, corrects the deviation with the help of the PID controller, and realizes closed-loop control to ensure precise operation; it accumulates operating data in a periodic manner, uses LSTM model training and optimization, explores the working condition adjustment rules, reversely iterates and optimizes the strategy library and adaptive algorithm parameters, and improves the system's adaptability to the complex environment of the ship.

2. The intelligent adjustment system for the adaptive multi-environment ship electric control valve actuator according to claim 1 is characterized in that: The execution process of the environment judgment module is as follows: It is spread all over the electric control valve and surrounding points to build a multi-dimensional sensing system, including: the temperature sensor uses a platinum resistance temperature detector; the vibration sensor uses a micro-electromechanical system accelerometer; electromagnetic interference monitoring uses a wide-band loop antenna with a spectrum analyzer; The data collected by each sensor is amplified, filtered, and converted into analogues by the signal conditioning circuit, and then input into a microprocessor with a built-in algorithm. Using big data cluster analysis and convolutional neural network algorithms in deep learning, we analyze short-term, medium-term, and long-term environmental data, train convolutional neural network models with typical ship operating condition samples, identify the current operating condition classification, and anchor the basic situation for subsequent adjustments; Implement differentiated monitoring paths to adapt to complex ship structures; optimize sensor layout according to the characteristics of different locations.

3. The intelligent adjustment system for the adaptive multi-environment ship electric control valve actuator according to claim 2 is characterized in that: The environmental judgment module implements the differentiated monitoring path as follows: Each electric control valve is given a unique identity code, and the coding rules are integrated into the position information. The sensor data is automatically classified and stored according to the code, and an independent data set is constructed, so that the data of the actuators at each position are in their place; when classifying electric control valves at different positions, they are distinguished by analyzing the similarity of the electric control valve monitoring data. The specific process is as follows: Clean the monitoring data collected from each electric control valve actuator; use statistical methods to identify and eliminate outliers; at the same time, check the integrity of the data, and use linear interpolation or mean filling methods to supplement missing data points; and normalize the dimensions and value ranges of different parameters through the minimum-maximum normalization method; Each electric control valve is regarded as a separate class. By calculating the cosine similarity between each pair of electric control valves, a pair of electric control valves with the highest cosine similarity is selected and merged into a new class. This process is repeated, and classes that meet the similarity standard are continuously merged until the preset stop condition is reached. The preset stop condition includes that the number of classes reaches a preset value or the similarity between classes is lower than a preset threshold, so as to obtain a hierarchical classification result, from the finest granularity of each electric control valve as a class to the final merging into several large categories; a unified control strategy is implemented for electric control valves of different categories.

4. The intelligent adjustment system for the adaptive multi-environment ship electric control valve actuator according to claim 1 is characterized in that: The specific operation steps of the adaptive adjustment module are as follows: Strategy library pre-storage and intelligent matching: Relying on the storage unit, the electric control valve regulation strategy library under the full spectrum of ship working conditions is built-in, and the motor drive voltage and current curves, valve opening adjustment step length, frequency, and speed limit parameter sets are stored according to the working conditions; the regulation strategy library is continuously optimized; Identify the working conditions based on the environmental judgment module and match the initial strategy blueprint; Dynamic fine-tuning mechanism: Fuzzy logic and particle swarm optimization algorithm are integrated to dynamically input the data of real-time monitoring module. Fuzzy logic quantifies the environmental and operational fuzzy quantities into control weights. The particle swarm optimization algorithm fine-tunes the strategy parameters in real time to optimize the target drive, and collaboratively generates the optimal adjustment instructions. The power drive unit controls the operation of the actuator motor to achieve smart and adaptive adjustment of the valve.

5. The intelligent adjustment system for the adaptive multi-environment ship electric control valve actuator according to claim 4 is characterized in that: The specific operation steps for continuously optimizing the adjustment strategy library in the adaptive adjustment module are as follows: During the operation of the ship, the actual operation data of the electric control valve actuator is continuously collected, including environmental parameters, actuator operation parameters and adjustment effect related data; the collected data is pre-processed to remove abnormal values ​​and noise interference; At the same time, the data are normalized so that the data of different parameters are at the same level; Regularly evaluate the performance of the electric control valve actuator and calculate the deviation between the actual operating data and the ideal performance indicators based on the preset performance indicators; Analyze the causes of the deviation and determine whether it is caused by environmental changes, actuator aging, and unreasonable control strategy; find out the key factors related to the deviation through statistical analysis of operating data; According to the results of performance evaluation and deviation analysis, the parameters in the regulation strategy library are adjusted in a targeted manner; For motor drive parameters, optimize them according to the actual operating status of the motor; Use time series analysis methods to monitor and predict the changing trends of temperature, humidity, vibration and electromagnetic interference environmental parameters, and combine the ship's navigation route and seasonal changes to predict extreme environmental conditions in advance; According to the environmental trend prediction results, the relevant strategies in the adjustment strategy library are pre-adjusted before the actual environmental changes; After the strategy is pre-adjusted, the new strategy is verified using the simulation environment. By establishing a mathematical model of the electric control valve and simulating the operation of the electric control valve under different environmental conditions, the effectiveness and feasibility of the pre-adjustment strategy are evaluated. When the actual environment changes and the pre-adjustment strategy is verified, the strategy parameters in the adjustment strategy library are updated online in real time to ensure that the actuator control system can stably switch to the new control strategy; during the switching process, the operating status of the actuator is monitored to prevent instability caused by strategy switching; Collect the operating data of the electric control valve actuator under different working conditions, including normal operation data and fault status data, as training samples for the machine learning model; divide these data into training set, validation set and test set, select machine learning algorithms, including neural network, support vector machine or decision tree, and build a performance prediction model for the electric control valve actuator; The model is trained with environmental parameters and actuator operating parameters as input features and valve control effect indicators as output targets. The optimal control strategy under different working conditions is predicted and optimized through the trained machine learning model. The model outputs the recommended motor drive voltage and current curves, valve opening adjustment step, frequency, and speed limit parameters based on the real-time input environment and operating data. Using reinforcement learning algorithms, the electric control valve actuator is allowed to explore and learn the optimal control strategy in actual operation; by setting a reward function, the actuator is encouraged to optimize performance indicators while meeting control requirements; As the ship’s operating time increases and new data accumulates, the machine learning model is updated regularly to adapt to changing operating conditions and actuator performance. Model fusion technology is used to fuse the prediction results of multiple different machine learning models to obtain better control strategy recommendations.

6. The intelligent adjustment system for the adaptive multi-environment ship electric control valve actuator according to claim 5 is characterized in that: The specific operation steps for evaluating the effectiveness and feasibility of the pre-adjustment strategy in the adaptive adjustment module are as follows: Comprehensive analysis is performed through the evaluation parameters of the pre-adjustment strategy, including: Valve opening error: Calculate the average value gt and standard deviation sd of the difference between the actual opening and the target opening of the electric control valve under simulated different environmental conditions; Flow control error: By establishing a mathematical relationship model between flow and valve opening, the theoretical flow is calculated according to the simulated valve opening, and compared with the actual simulated flow value to obtain the flow control error wf; Response time: record the time from receiving the control signal to the time when the electric control valve reaches the target opening of 90%; when simulating different environmental conditions, compare the changes in response time and calculate the average response time he; Overshoot: Calculate the maximum deviation of the electric control valve exceeding the target opening during the adjustment process and the ratio of the target opening kb, expressed as a percentage; The obtained average opening difference gt, standard deviation of opening difference sd, mean response time he and opening ratio kb are normalized and entered into the following formula: To obtain the comprehensive judgment value FD, where They are the preset weight coefficients of the opening difference average value gt, the opening difference standard deviation sd, the response time mean value he and the opening ratio kb, and the obtained comprehensive judgment value FD is used as the criterion for measuring the effectiveness and feasibility of the adjustment strategy; when the comprehensive judgment value FD does not exceed the preset comprehensive judgment threshold, the effectiveness and feasibility of the pre-adjustment strategy are judged to meet the standards.

7. The intelligent adjustment system for the adaptive multi-environment ship electric control valve actuator according to claim 1 is characterized in that: The execution process of the feedback optimization module is as follows: Real-time deviation correction: Compare the actual operating parameters of the actuator with the preset targets of the adjustment instructions. With the help of the proportional-integral-differential controller, once the deviation exceeds the allowable threshold, the deviation correction amount is calculated in real time according to the proportional-integral-differential algorithm, and the adjustment instructions are corrected and sent back to the adaptive adjustment module. Closed-loop control ensures accurate tracking of the target. Self-learning evolution: accumulate operating data by voyage or monthly cycle, feed it to the long-short term memory network model training optimization of machine learning, explore the working condition-adjustment rules, reverse iterate and optimize the strategy library and adaptive algorithm parameters, improve the system's adaptability to the complex environment of the ship, extend the equipment life, and improve the efficiency of ship operation and maintenance.

Images