A tensioning device with intelligent frequency conversion and control method

Abstract

The present invention discloses an intelligent variable frequency tensioning device and control method. The device includes six units, including a multimodal data sensing and interactive integration unit and an adaptive variable frequency drive execution and control unit. The multimodal data sensing and interactive integration unit collects the tension, displacement, and speed signals of the tensioning device during operation. The hierarchical parameter optimization processing center extracts features and performs preliminary processing on the data. The intelligent algorithm collaborative decision-making module integrates the optimized PID control algorithm and RBF neural network model to generate a control strategy. The adaptive variable frequency drive execution and control unit drives the motor, driving the mechanical structure dynamic adjustment linkage unit to adjust the tensioning components. The state monitoring and feedback correction unit monitors and feeds back data in real time for strategy and parameter correction. The device and method realize intelligent variable frequency control of the tensioning device, effectively improving control accuracy and system stability, and meeting the requirements of complex working conditions.

Description

Technical Field

[0001] The present invention relates to the field of tensioning device control, and in particular to an intelligent variable frequency tensioning device and a control method. Background Art

[0002] With the acceleration of industrial automation, various continuous conveying systems, cable processing equipment, and textile machinery are placing increasingly stringent performance requirements on tensioning devices. As a critical component for ensuring stable equipment operation, improving production efficiency, and enhancing product quality, the performance of tensioning devices directly impacts the reliability and stability of the entire system. In the complex and ever-changing working conditions of modern industry, traditional tensioning devices are no longer able to meet the demands for high precision, high response speed, and adaptive adjustment. Intelligent and automated technology upgrades are urgently needed to adapt to diverse industrial application scenarios.

[0003] At present, traditional tensioning devices have significant deficiencies in control strategies and structural design. On the one hand, in terms of control strategies, most use fixed-parameter PID control algorithms, which lack the ability to adaptively adjust to complex working conditions and changes in system parameters. When the tensioning device faces sudden load changes, environmental interference, or mechanical wear caused by long-term operation, it is unable to adjust the control parameters in a timely manner, which can easily lead to problems such as adjustment lag and large overshoot, making it difficult to achieve high-precision tension and speed control. On the other hand, in terms of structural design, the mechanical structure of traditional tensioning devices has poor coordination with the drive system, and the linkage response speed between the various components is slow, making it impossible to quickly adapt to changes in working conditions. In addition, its monitoring and feedback mechanism is imperfect, and it cannot accurately obtain the operating status information of the tensioning device in real time, and cannot provide an effective basis for the optimization and adjustment of the control strategy. This leads to low overall operating efficiency and poor stability of the device, which seriously restricts the development of intelligent industrial production. Summary of the Invention

[0004] In order to overcome the shortcomings and deficiencies of the prior art, the present invention provides an intelligent variable frequency tensioning device and a control method.

[0005] The technical solution adopted by the present invention is an intelligent variable frequency tensioning device, including: a multimodal data sensing and interaction integration unit, which establishes a two-way data transmission link with an external tension sensor, a displacement sensor, and a speed sensor through a high-speed serial bus, and is used to collect real-time tension signals, displacement signals, and speed signals during the operation of the tensioning device; an adaptive variable frequency drive execution control unit, which is connected to the multimodal data sensing and interaction integration unit through an industrial Ethernet protocol, receives the data signal transmitted by the sensing unit, and performs variable frequency speed control on the drive motor of the tensioning device based on preset variable frequency drive parameters; a hierarchical parameter optimization processing center, whose input port is connected to the data output end of the multimodal data sensing and interaction integration unit, and the output port is respectively connected to the data input end of the adaptive variable frequency drive execution control unit and the intelligent algorithm collaborative decision-making module, to process the collected multimodal data The data is extracted and preliminarily processed; the intelligent algorithm collaborative decision-making module establishes data communication channels with the hierarchical parameter optimization processing center and the adaptive variable frequency drive execution control unit respectively, and has built-in optimized PID control algorithm and RBF neural network model, and calculates and makes decisions on the control strategy based on the processed data; the mechanical structure dynamic adjustment linkage unit, whose control signal input end is connected to the control signal output end of the adaptive variable frequency drive execution control unit, is connected to the tensioning component of the tensioning device through a transmission mechanism, and dynamically adjusts the tensioning component according to the received control signal; the state monitoring and feedback correction unit establishes data interaction links with the multimodal data perception and interaction integration unit, the hierarchical parameter optimization processing center, and the intelligent algorithm collaborative decision-making module respectively, monitors the operating status of the tensioning device in real time, and feeds back the monitoring data to the corresponding module for parameter correction and control strategy adjustment.

[0006] Furthermore, in the intelligent algorithm collaborative decision-making module, the tension control model constructed based on the optimized PID control algorithm and RBF neural network is:

[0007]

[0008] in, The target tension value output by the tensioning device; is the proportional coefficient, which is set according to the load characteristics and mechanical structure parameters of the tensioning device; is the deviation between the actual tension value and the target tension value at the current moment; is the integral time constant, which is determined by the response speed requirement of the tensioning device; is the differential time constant, which is adjusted based on the dynamic characteristics of the tensioner; is the number of nodes in the hidden layer of the RBF neural network, which is set according to the control accuracy requirements of the tensioning device; The hidden layer The connection weight from each node to the output layer; is the radial basis function, is the input vector, including real-time collected tension, displacement, and velocity data. For the The center vector of hidden layer nodes, It is the deviation between the actual tension value and the target tension value.

[0009] Furthermore, in the intelligent algorithm collaborative decision-making module, the speed regulation model constructed based on the optimized PID control algorithm and RBF neural network is:

[0010]

[0011] in, The adjustment speed value of the tensioning device drive motor; is the proportional coefficient, which is determined according to the power parameters of the drive motor and the transmission ratio of the tensioning device; is the deviation between the actual speed value and the target speed value at the current moment; is the integral time constant, which is set according to the starting characteristics of the drive motor; is the differential time constant, which is adjusted by the speed regulation accuracy requirement of the drive motor; is the number of nodes in the hidden layer of the RBF neural network, which is set based on the dynamic response requirements of speed control; The hidden layer The connection weight from each node to the output layer; is the radial basis function, is the input vector, including the speed and load torque data collected in real time, For the The center vector of hidden layer nodes, is the deviation between the actual speed value and the target speed value.

[0012] Furthermore, in the hierarchical parameter optimization processing center, the model for extracting features from multimodal data is:

[0013] in, is the extracted feature data vector; is the original tension data, A feature extraction function for tension data is constructed based on the sampling frequency and accuracy parameters of the tension sensor of the tensioning device; is the original displacement data, To extract the feature function of displacement data, it is set according to the travel range and resolution parameters of the tensioning device; is the original speed data, It is a feature extraction function for speed data, which is determined according to the speed range of the drive motor and the encoder accuracy parameters; Represents the fusion operation of feature data.

[0014] Furthermore, in the adaptive variable frequency drive execution control unit, the drive current regulation model constructed based on the optimized PID control algorithm and RBF neural network is:

[0015]

[0016] in, Adjust the current value for driving the motor; is the proportional coefficient, which is set according to the rated current and power factor parameters of the drive motor; is the deviation between the actual current value and the target current value at the current moment; is the integration time constant, which is determined by the overload capacity and heat dissipation characteristics of the drive motor; is the differential time constant, which is adjusted based on the response speed parameters of the drive circuit; is the number of hidden layer nodes of the RBF neural network, which is set according to the stability requirements of the drive current control; The hidden layer The connection weight from each node to the output layer; is the radial basis function, is the input vector, including the current, voltage, and load torque data collected in real time. For the The center vector of hidden layer nodes, is the deviation between the actual current value and the target current value.

[0017] Furthermore, in the state monitoring and feedback correction unit, the abnormal state warning model constructed based on the optimized PID control algorithm and RBF neural network is:

[0018]

[0019] in, It is an abnormal status flag. When the value is 1, it indicates an abnormality, and when the value is 0, it indicates normal. To monitor the real-time tension data, It is an abnormality judgment function based on tension data, which is constructed according to the rated tension and safety factor parameters of the tensioning device; is the threshold of abnormal tension; To monitor the real-time displacement data, It is an abnormality judgment function based on displacement data and is set according to the limit stroke parameters of the tensioning device; is the displacement anomaly threshold; To monitor real-time speed data, It is an abnormality judgment function based on speed data, which is determined according to the rated speed and speed regulation range parameters of the drive motor; is the speed abnormality threshold.

[0020] Furthermore, the hierarchical parameter optimization processing center includes a data preprocessing submodule, which filters the collected original multimodal data and adopts a median filtering algorithm based on a sliding window to remove noise interference in the data; a feature extraction submodule, which uses a feature extraction method based on local mean decomposition to decompose the multimodal data into multiple feature components, and extracts calibration feature parameters that can reflect the operating status of the tensioning device; a parameter optimization submodule, which performs global optimization on the parameters of the optimized PID control algorithm and RBF neural network based on a genetic algorithm, and continuously adjusts the parameter combination through selection, crossover, and mutation operations to adapt to the operating requirements of the tensioning device under different working conditions; a data cache submodule, which adopts a double buffering mechanism to temporarily store the data before and after processing to ensure the continuity and stability of data processing and prevent data loss or processing interruption.

[0021] Furthermore, the intelligent algorithm collaborative decision-making module includes a PID parameter adaptive adjustment unit, which monitors the operating status of the tensioning device in real time, and dynamically adjusts the proportional, integral, and differential parameters through the optimized PID control algorithm according to the collected data, so that the tensioning device responds quickly and reaches a stable operating state; an RBF neural network training unit, which uses historical operating data to perform offline training on the RBF neural network, and improves the prediction accuracy and generalization ability of the neural network by adjusting the center vector, width parameter, and connection weight of the hidden layer node; a control strategy fusion unit, which performs weighted fusion on the control quantity output by the optimized PID control algorithm and the control quantity predicted by the RBF neural network, sets the weight coefficient according to different working conditions, and generates the final control strategy; a decision result output unit, which converts the generated control strategy into a standard control signal format and transmits it to the adaptive variable frequency drive execution control unit through a data communication link.

[0022] Furthermore, the adaptive variable frequency drive execution control unit includes a drive signal generation module, which generates a corresponding PWM drive signal according to the received control strategy to control the speed and torque of the drive motor; a power amplification module, which amplifies the generated PWM drive signal to drive the normal operation of the drive motor of the tensioning device; a motor protection module, which monitors the current, voltage, and temperature parameters of the drive motor in real time, and immediately cuts off the motor power supply when an abnormality is detected to prevent damage to the motor; a drive parameter configuration module, which allows the user to manually configure the rated speed, rated torque, and acceleration and deceleration time parameters of the drive motor according to different application scenarios and load requirements of the tensioning device.

[0023] A control method for an intelligent variable frequency tensioning device comprises the following steps:

[0024] Step S1: The multimodal data sensing and interaction integration unit collects the tension signal, displacement signal, and speed signal of the tensioning device in real time during operation according to a preset sampling frequency and communication protocol, and transmits the collected data to the hierarchical parameter optimization processing center via a high-speed serial bus;

[0025] Step S2: The hierarchical parameter optimization processing center performs data cleaning and feature extraction processing on the received data, uses a preset data processing algorithm to remove interference components in the data, extracts calibration feature parameters that can characterize the operating status of the tensioning device, and transmits the processed data to the intelligent algorithm collaborative decision module;

[0026] Step S3: The intelligent algorithm collaborative decision-making module calls the built-in optimized PID control algorithm and RBF neural network model based on the received processed data to calculate and make decisions on the control strategy of the tensioning device. In combination with various parameters of the tensioning device, the algorithm generates control instructions for the drive motor through iterative calculation;

[0027] Step S4: The control instruction generated by the intelligent algorithm collaborative decision module is transmitted to the adaptive variable frequency drive execution control unit, which converts the control instruction into a corresponding drive signal, and after power amplification processing, drives the drive motor of the tensioning device to operate, thereby driving the mechanical structure dynamic adjustment linkage unit to adjust the tensioning component;

[0028] Step S5: The state monitoring and feedback correction unit monitors the operating state of the tensioning device in real time, collects various state data during operation, and feeds the data back to the hierarchical parameter optimization processing center and the intelligent algorithm collaborative decision-making module for real-time correction of the control strategy and algorithm parameters;

[0029] Step S6: Repeat the above steps of data collection, processing, decision-making, execution and feedback correction to perform intelligent frequency conversion control and continuous stable operation of the tensioning device.

[0030] Beneficial Effects: The present invention proposes an intelligent variable-frequency tensioning device and control method. To address the lack of adaptive adjustment capabilities in traditional control strategies, the device's intelligent algorithm collaborative decision-making module incorporates an optimized PID control algorithm and an RBF neural network model, which work in tandem. The optimized PID control algorithm dynamically adjusts proportional, integral, and differential parameters based on real-time operating conditions. The RBF neural network accurately predicts control parameters under different operating conditions by learning from extensive historical data. This enables the device to rapidly respond and achieve high-precision tension and speed control even under complex conditions such as sudden load changes and environmental interference, avoiding adjustment lag and overshoot. To address the shortcomings of traditional devices, which suffer from poor coordination between the mechanical structure and drive system and imperfect monitoring and feedback mechanisms, a multimodal data perception and interaction integration unit collects multi-dimensional data such as tension, displacement, and speed in real time. A hierarchical parameter optimization processing center performs in-depth feature extraction and preliminary processing on the data, and a state monitoring and feedback correction unit monitors the device's operating status in real time and provides feedback. Furthermore, the adaptive variable-frequency drive execution and control unit works closely with the mechanical structure dynamic adjustment linkage unit to precisely drive the motor and adjust the tensioning components according to control commands, ensuring efficient linkage between all components. In addition, each module achieves fast and stable data interaction through specific data transmission protocols and communication links, enabling the device to obtain operating status information in real time, providing a reliable basis for optimizing and adjusting the control strategy, greatly improving overall operating efficiency and stability, and meeting the modern industry's demand for high-precision and intelligent application of tensioning devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a composition diagram of the device unit of the present invention;

[0032] Figure 2 The figure is a flow chart of the method steps of the present invention. DETAILED DESCRIPTION

[0033] It should be noted that, unless there is a conflict, the embodiments in this application and the features in the embodiments can be combined with each other. The application is further described in detail below with reference to the drawings and specific embodiments.

[0034] like Figure 1 As shown, an intelligent variable frequency tensioning device includes:

[0035] A multimodal data sensing and interaction integrated unit, which establishes a bidirectional data transmission link with external tension sensors, displacement sensors, and speed sensors via a high-speed serial bus to collect real-time tension signals, displacement signals, and speed signals during the operation of the tensioning device;

[0036] Specifically, the multimodal data perception and interaction integration unit is the key entry point for the entire intelligent variable frequency tensioning device to obtain external information. The unit establishes a two-way data transmission link with the external tension sensor, displacement sensor, and speed sensor through a high-speed serial bus. In terms of technical parameters, the high-speed serial bus has the characteristics of high transmission rate and strong anti-interference ability. Its data transmission rate can reach hundreds of megabytes, which can ensure that the real-time data collected by the sensor is transmitted quickly and stably. The tension sensor uses a high-precision strain gauge sensor with a measurement accuracy of up to , accurately sensing tension changes during tensioning device operation. The displacement sensor uses a laser displacement sensor with micron-level resolution, accurately measuring the displacement of the tensioning component. The speed sensor uses an incremental photoelectric encoder, generating thousands of pulses per revolution, accurately measuring the device's operating speed. This unit periodically collects sensor data at a preset sampling frequency, enabling real-time monitoring of the tensioning device's operating status.

[0037] During implementation, the multimodal data perception and interaction integration unit has built-in data acquisition modules and communication protocol processing modules. The data acquisition module uses corresponding signal conditioning circuits based on the output signal types of different sensors to amplify, filter, and process the weak electrical signals output by the sensors, converting them into digital signals suitable for acquisition. The communication protocol processing module follows standard industrial communication protocols, such as the Modbus RTU protocol, to encapsulate the collected data so that it can be accurately transmitted on a high-speed serial bus. The significance of this unit is to provide reliable operating data for the entire tensioning device and is the basis for realizing intelligent control. Only by accurately acquiring multimodal data such as tension, displacement, and speed can subsequent control and decision-making be carried out effectively.

[0038] The adaptive variable frequency drive execution control unit is connected to the multimodal data perception and interaction integration unit via the industrial Ethernet protocol, receives the data signal transmitted by the perception unit, and performs variable frequency speed control on the drive motor of the tensioning device based on the preset variable frequency drive parameters;

[0039] Specifically, the adaptive variable frequency drive execution and control unit is responsible for converting control signals into actual drive actions. It connects to the multimodal data perception and interaction integrated unit via the Industrial Ethernet protocol. Industrial Ethernet offers fast data transmission rates and excellent compatibility, with transmission rates reaching up to 1 Gbps. This unit receives operational data signals from the perception unit in real time. This unit incorporates a high-performance variable frequency controller, whose core control chip utilizes a dedicated digital signal processor (DSP) capable of rapidly processing complex control algorithms. The variable frequency controller precisely regulates the tensioner's drive motor based on preset variable frequency drive parameters, such as rated power, rated speed, and rated torque. By adjusting the motor's supply frequency, the motor speed can be continuously adjusted over a wide range, from 0 to 120% of the rated speed, to meet the tensioner's operating speed requirements under varying operating conditions.

[0040] During implementation, the adaptive variable frequency drive execution and control unit includes a drive signal generation module, a power amplification module, and a motor protection module. The drive signal generation module utilizes pulse-width modulation (PWM) technology to generate the corresponding drive signal based on the control instructions received. The motor's speed and torque are controlled by adjusting the PWM signal's duty cycle. The power amplification module utilizes high-power IGBT power devices to amplify the drive signal, ensuring sufficient drive power to operate the motor. The motor protection module monitors motor parameters such as current, voltage, and temperature in real time. Upon detecting an abnormality such as overcurrent, overvoltage, or overheating, the protection mechanism is immediately triggered, cutting off the motor's power supply to prevent damage. This unit is a key executive component for implementing intelligent variable frequency control of the tensioning device. Through precise control of the drive motor, it ensures stable and efficient operation of the tensioning device.

[0041] The hierarchical parameter optimization processing center has its input port connected to the data output port of the multimodal data perception and interaction integration unit, and its output port is connected to the data input ports of the adaptive variable frequency drive execution control unit and the intelligent algorithm collaborative decision-making module respectively, to perform feature extraction and preliminary processing on the collected multimodal data;

[0042] Specifically, the hierarchical parameter optimization processing center plays a role in data processing and preliminary decision-making in the intelligent variable frequency tensioning device. Its input port is connected to the data output port of the multimodal data perception and interaction integration unit to receive the collected original multimodal data. The center adopts a multi-level processing architecture, including a data preprocessing layer, a feature extraction layer, and a parameter optimization layer. In the data preprocessing layer, digital filtering technology is used to process the raw data to remove noise and interference signals in the data, thereby improving the accuracy and reliability of the data. In the feature extraction layer, advanced signal processing algorithms, such as the empirical mode decomposition (EMD) algorithm, are used to extract key characteristic parameters that can reflect the operating status of the tensioning device from the preprocessed data, such as tension fluctuation characteristics, displacement change trend characteristics, etc. In the parameter optimization layer, the parameters of the subsequent control algorithm are preliminarily optimized based on the optimization algorithm to provide high-quality input data for the intelligent algorithm collaborative decision-making module.

[0043] During implementation, the hierarchical parameter optimization processing center uses a high-performance microprocessor as the core processing unit, which has strong data processing capabilities and computing speed. Each processing layer implements the corresponding functional module through software programming. The digital filtering algorithm of the data preprocessing layer uses FIR filters to set appropriate filter coefficients according to the noise characteristics of the data; the EMD algorithm of the feature extraction layer obtains the characteristic components of different frequency components by multiple screening and decomposition of the data; the parameter optimization layer uses a genetic algorithm to simulate the selection, crossover, mutation and other operations in the biological evolution process to perform global optimization of the control algorithm parameters. The significance of this center lies in the in-depth processing and analysis of the raw data, extracting valuable information, providing reliable data support for subsequent intelligent control, and improving the control accuracy and response speed of the entire device.

[0044] The intelligent algorithm collaborative decision-making module establishes data communication channels with the hierarchical parameter optimization processing center and the adaptive variable frequency drive execution control unit. It has a built-in optimized PID control algorithm and RBF neural network model to calculate and make control strategies based on processed data.

[0045] Specifically, the intelligent algorithm collaborative decision-making module is the core decision-making unit of the entire intelligent variable frequency tensioning device. It establishes data communication channels with the hierarchical parameter optimization processing center and the adaptive variable frequency drive execution control unit, receives processed multimodal data, and outputs control strategies. The module has built-in optimized PID control algorithm and RBF neural network model, which work together to give full play to their respective advantages. The optimized PID control algorithm can adjust the proportional, integral, and differential parameters in real time according to the current operating status of the system to achieve rapid response and stable control of the system. The RBF neural network model has powerful nonlinear mapping capabilities and can accurately predict the optimal control parameters of the system under different working conditions by learning from a large amount of historical data. By combining the two algorithms, the module can generate the optimal control strategy for the complex and changeable operating conditions of the tensioning device, ensuring that the device can achieve high-precision tension and speed control under various working conditions.

[0046] During implementation, the intelligent algorithm collaborative decision-making module uses a high-performance multi-core processor to provide powerful computing power support for the operation of the algorithm. The optimized PID control algorithm uses software programming to achieve adaptive parameter adjustment and calculate the current control quantity based on real-time collected data. The RBF neural network model uses a large amount of historical operating data through offline training to adjust and optimize the network parameters. After training, the network model is stored in the module's memory. During actual operation, the collected data is input into the network model to obtain the predicted control parameters. Finally, the algorithm fusion module performs a weighted fusion of the control quantities output by the optimized PID control algorithm and the RBF neural network model. Appropriate weight coefficients are set according to different working conditions to generate the final control strategy, which is then transmitted to the adaptive variable frequency drive execution control unit. The significance of this module is to achieve intelligent control of the tensioning device, improve the device's adaptability and control accuracy, and enable it to adapt to various complex industrial application scenarios.

[0047] The mechanical structure dynamic adjustment linkage unit has a control signal input terminal connected to the control signal output terminal of the adaptive variable frequency drive execution control unit, is connected to the tensioning component of the tensioning device through a transmission mechanism, and dynamically adjusts the tensioning component according to the received control signal;

[0048] Specifically, the mechanical structure dynamic adjustment linkage unit is the key unit that converts control signals into actual mechanical actions. Its control signal input is connected to the control signal output of the adaptive variable frequency drive execution and control unit, receiving control commands from the drive unit. This unit primarily consists of a transmission mechanism and a tensioning component. The transmission mechanism utilizes a combination of high-precision gear transmission and ball screw transmission, characterized by high transmission efficiency and precision. The gear transmission has a precise transmission ratio, accurately transmitting the speed and torque of the drive motor to the tensioning component. The ball screw transmission utilizes rolling friction between the nut and the screw, resulting in a low friction coefficient, a transmission efficiency exceeding 90%, and high positioning accuracy, enabling precise displacement control of the tensioning component. Depending on the application scenario, the tensioning component can take the form of a hydraulic cylinder, pneumatic cylinder, or electric push rod, capable of generating sufficient tension to meet the operating requirements of the tensioning device.

[0049] During implementation, the transmission mechanism and tensioning components of the mechanical structure's dynamic adjustment linkage unit are tightly coupled through mechanical connections. When the adaptive variable frequency drive execution control unit issues a control command, the drive motor begins operating, transmitting power to the ball screw via gear transmission. The ball screw converts rotational motion into linear motion, driving the tensioning component to adjust its displacement, thereby achieving dynamic adjustment of the tensioning device's tension. During the adjustment process, the displacement sensor installed on the tensioning component provides real-time feedback on the tensioning component's position, forming a closed-loop control loop that ensures the tensioning component accurately reaches the target position and achieves precise tension control. The significance of this unit lies in achieving close coordination between the tensioning device's mechanical structure and control system, enabling the device to adjust the tensioning state in real time according to the control strategy, ensuring stable system operation.

[0050] The status monitoring and feedback correction unit establishes data interaction links with the multimodal data perception and interaction integration unit, the hierarchical parameter optimization processing center, and the intelligent algorithm collaborative decision-making module, respectively, to monitor the operating status of the tensioning device in real time, and feed back the monitoring data to the corresponding module for parameter correction and control strategy adjustment.

[0051] Specifically, the status monitoring and feedback correction unit is responsible for all-round real-time monitoring of the operating status of the intelligent variable frequency tensioning device, and feeding back the monitoring data to the relevant modules to achieve precise control and optimization of the operating status of the device. This unit establishes data interaction links with the multimodal data perception and interaction integration unit, the hierarchical parameter optimization processing center, and the intelligent algorithm collaborative decision-making module. It not only collects basic data such as tension, displacement, and speed from the sensor in real time, but also monitors the electrical parameters of the device, such as the current, voltage, and power factor of the drive motor, as well as the temperature, vibration and other status information of the mechanical components. Through real-time analysis of these multi-dimensional data, abnormal conditions that occur during the operation of the device can be discovered in a timely manner, such as sudden changes in tension, motor overload, and component overheating.

[0052] During implementation, the condition monitoring and feedback correction unit utilizes a distributed data acquisition architecture, collecting different types of condition data through multiple data acquisition modules. These data acquisition modules are highly precise and reliable, capable of accurately capturing a wide range of weak signals. After signal conditioning and analog-to-digital conversion, the collected data is transmitted to the central processing unit via a data communications network. The central processing unit uses pre-set data analysis algorithms and fault diagnosis models to conduct in-depth analysis and processing of the data. Upon detecting an abnormality, a corresponding feedback signal is immediately generated and transmitted to the hierarchical parameter optimization processing center and the intelligent algorithm collaborative decision-making module. The hierarchical parameter optimization processing center adjusts the data processing parameters based on the feedback signal, while the intelligent algorithm collaborative decision-making module corrects the control strategy and algorithm parameters based on the feedback information, thereby dynamically optimizing and adjusting the tensioning device's operating status. This unit is crucial for ensuring the stable and reliable operation of the tensioning device. Through real-time monitoring and feedback correction, it can effectively prevent failures and improve the device's operating efficiency and service life.

[0053] Preferably, in the intelligent algorithm collaborative decision-making module, the tension control model constructed based on the optimized PID control algorithm and RBF neural network is:

[0054]

[0055] in, The target tension value output by the tensioning device; is the proportional coefficient, which is set according to the load characteristics and mechanical structure parameters of the tensioning device; is the deviation between the actual tension value and the target tension value at the current moment; is the integral time constant, which is determined by the response speed requirement of the tensioning device; is the differential time constant, which is adjusted based on the dynamic characteristics of the tensioner; is the number of nodes in the hidden layer of the RBF neural network, which is set according to the control accuracy requirements of the tensioning device; The hidden layer The connection weight from each node to the output layer; is the radial basis function, is the input vector, including real-time collected tension, displacement, and velocity data. For the The center vector of hidden layer nodes, It is the deviation between the actual tension value and the target tension value.

[0056] Specifically, the construction and application of a tension control model within the intelligent algorithm collaborative decision-making module deeply integrates an optimized PID control algorithm with an RBF neural network, aiming to achieve precise control of the target tension output by the tensioning device. The optimized PID control algorithm sets a proportional coefficient based on the tensioning device's load characteristics and mechanical structure parameters, determines the integral time constant according to the device's response speed requirements, and adjusts the differential time constant based on its dynamic characteristics. This allows for rapid and stable response to deviations between actual and target tensions. The RBF neural network component sets the number of hidden layer nodes based on control accuracy requirements. By processing input vectors containing real-time tension, displacement, and velocity data, and utilizing the connection weights and radial basis functions from hidden layer nodes to the output layer, the model explores tension variation patterns under complex working conditions. The model is implemented through a software program running within the intelligent algorithm collaborative decision-making module. It first collects real-time data transmitted by the multimodal data perception and interaction integration unit, processes it in a hierarchical parameter optimization processing center, and then inputs it into the tension control model. The model calculates the target tension value and transmits the result to the adaptive variable frequency drive execution control unit, which drives the motor to adjust the tensioning components to form a closed-loop control, effectively improving the accuracy and stability of the tension control of the tensioning device and ensuring that the tension output meets the requirements under different working conditions.

[0057] Preferably, in the intelligent algorithm collaborative decision-making module, the speed regulation model constructed based on the optimized PID control algorithm and RBF neural network is:

[0058]

[0059] in, The adjustment speed value of the tensioning device drive motor; is the proportional coefficient, which is determined according to the power parameters of the drive motor and the transmission ratio of the tensioning device; is the deviation between the actual speed value and the target speed value at the current moment; is the integral time constant, which is set according to the starting characteristics of the drive motor; is the differential time constant, which is adjusted by the speed regulation accuracy requirement of the drive motor; is the number of nodes in the hidden layer of the RBF neural network, which is set based on the dynamic response requirements of speed control; The hidden layer The connection weight from each node to the output layer; is the radial basis function, is the input vector, including the speed and load torque data collected in real time, For the The center vector of hidden layer nodes, is the deviation between the actual speed value and the target speed value.

[0060] Specifically, the speed regulation model within the intelligent algorithm collaborative decision-making module also integrates an optimized PID control algorithm and a RBF neural network to precisely regulate the speed of the tensioner drive motor. The optimized PID control algorithm sets a proportional coefficient based on the drive motor power parameters and the tensioner transmission ratio, determines the integral time constant according to the motor's starting characteristics, and adjusts the differential time constant based on the required speed regulation accuracy to effectively regulate the deviation between the actual and target motor speeds. The RBF neural network sets the number of hidden layer nodes based on the dynamic response requirements of speed control. It analyzes the input vector containing real-time speed and load torque data and uses the connection weights from the hidden layer nodes to the output layer and radial basis functions to predict the appropriate motor speed under different operating conditions. During actual operation, speed and other data collected by the multimodal data perception and interaction integration unit are processed by the hierarchical parameter optimization processing center and then input into the speed regulation model. The model calculates the adjusted speed value and transmits it to the adaptive variable frequency drive execution and control unit, which generates a corresponding drive signal to control the motor speed, ensuring that the tensioner maintains a stable and accurate operating speed under varying loads and operating conditions, thereby improving the overall operational efficiency and reliability of the device.

[0061] Preferably, in the hierarchical parameter optimization processing center, the model for extracting features from multimodal data is: ,in, is the extracted feature data vector; is the original tension data, A feature extraction function for tension data is constructed based on the sampling frequency and accuracy parameters of the tension sensor of the tensioning device; is the original displacement data, To extract the feature function of displacement data, it is set according to the travel range and resolution parameters of the tensioning device; is the original speed data, It is a feature extraction function for speed data, which is determined according to the speed range of the drive motor and the encoder accuracy parameters; Represents the fusion operation of feature data.

[0062] Specifically, the multimodal data feature extraction model within the hierarchical parameter optimization processing hub constructs feature extraction functions for the raw tension, displacement, and velocity data based on parameters such as the tensioner sensor's sampling frequency, accuracy, travel range, and resolution, as well as the drive motor's speed range and encoder accuracy. These functions process the raw data and integrate the processed feature data into feature data vectors using specific fusion operations. During implementation, the raw data collected by the multimodal data perception and interaction integration unit is transmitted to the hierarchical parameter optimization processing hub. The data preprocessing submodule removes noise interference before entering the feature extraction submodule. The feature extraction submodule utilizes the aforementioned model to extract key features from the raw data that reflect the tensioner's operating status, such as tension fluctuations, displacement trends, and velocity patterns. These feature data vectors provide the core basis for the subsequent intelligent algorithm collaborative decision-making module to formulate control strategies, helping to improve the control strategy's relevance and effectiveness, ensuring accurate assessment and control of the tensioner's operating status.

[0063] Preferably, in the adaptive variable frequency drive execution control unit, the drive current regulation model constructed based on the optimized PID control algorithm and RBF neural network is:

[0064]

[0065] in, Adjust the current value for driving the motor; is the proportional coefficient, which is set according to the rated current and power factor parameters of the drive motor; is the deviation between the actual current value and the target current value at the current moment; is the integration time constant, which is determined by the overload capacity and heat dissipation characteristics of the drive motor; is the differential time constant, which is adjusted based on the response speed parameters of the drive circuit; is the number of hidden layer nodes of the RBF neural network, which is set according to the stability requirements of the drive current control; The hidden layer The connection weight from each node to the output layer; is the radial basis function, is the input vector, including the current, voltage, and load torque data collected in real time. For the The center vector of hidden layer nodes, is the deviation between the actual current value and the target current value.

[0066] Specifically, the drive current regulation model in the adaptive variable frequency drive execution control unit combines an optimized PID control algorithm with an RBF neural network to achieve precise control of the drive motor's adjusted current value. The optimized PID control algorithm sets the proportional coefficient based on the drive motor's rated current and power factor parameters, determines the integral time constant based on the motor's overload capacity and heat dissipation characteristics, and adjusts the differential time constant based on the drive circuit's response speed parameters to effectively regulate the deviation between the motor's actual and target currents. The RBF neural network sets the number of hidden layer nodes based on the stability requirements of the drive current control, analyzes the input vector containing real-time current, voltage, and load torque data, and uses the connection weights and radial basis functions from the hidden layer nodes to the output layer to predict the appropriate drive current under different operating conditions. In actual operation, the relevant data collected by the multimodal data perception and interaction integration unit is processed by the hierarchical parameter optimization processing center and then input into the drive current regulation model. After the model calculates the adjusted current value, it is transmitted to the drive signal generation module of the adaptive variable frequency drive execution control unit to generate the corresponding PWM drive signal. After amplification by the power amplifier module, it drives the motor, ensuring that the motor can operate with appropriate current under different loads and working conditions, avoiding damage to the motor due to current abnormalities, and improving the stability and reliability of the drive system.

[0067] Preferably, in the state monitoring and feedback correction unit, the abnormal state early warning model constructed based on the optimized PID control algorithm and RBF neural network is:

[0068]

[0069] in, It is an abnormal status flag. When the value is 1, it indicates an abnormality, and when the value is 0, it indicates normal. To monitor the real-time tension data, It is an abnormality judgment function based on tension data, which is constructed according to the rated tension and safety factor parameters of the tensioning device; is the threshold of abnormal tension; To monitor the real-time displacement data, It is an abnormality judgment function based on displacement data and is set according to the limit stroke parameters of the tensioning device; is the displacement anomaly threshold; To monitor real-time speed data, It is an abnormality judgment function based on speed data, which is determined according to the rated speed and speed regulation range parameters of the drive motor; is the speed abnormality threshold.

[0070] Specifically, the abnormal state warning model in the state monitoring and feedback correction unit constructs an abnormality judgment function based on the tension, displacement, and speed monitoring data according to the rated tension, safety factor, limit stroke, and rated speed and speed regulation range of the tensioning device, and sets the corresponding abnormality threshold. In terms of implementation, the state monitoring and feedback correction unit collects real-time monitoring data transmitted by the multimodal data perception and interactive integration unit in real time, and substitutes the tension, displacement, and speed data into the corresponding abnormality judgment function for calculation. If the calculated result of any monitoring data exceeds the corresponding abnormality threshold, the abnormal state flag is set to 1, indicating that the device has an abnormality; otherwise, it is set to 0, indicating normal operation. Once an abnormality is detected, the unit immediately transmits the feedback signal to the hierarchical parameter optimization processing center and the intelligent algorithm collaborative decision-making module, prompting the two to adjust the control strategy and algorithm parameters, promptly eliminate potential fault hazards, ensure the safe and stable operation of the tensioning device, and avoid equipment damage or production accidents caused by abnormal operation.

[0071] Preferably, the hierarchical parameter optimization processing center includes a data preprocessing submodule, which performs filtering on the collected original multimodal data, adopts a median filtering algorithm based on a sliding window to remove noise interference in the data, improve the accuracy and reliability of the data, and provide a high-quality data basis for subsequent feature extraction and parameter optimization; a feature extraction submodule, which uses a feature extraction method based on local mean decomposition to decompose the multimodal data into multiple feature components, and extracts calibration feature parameters that can reflect the operating status of the tensioning device; a parameter optimization submodule, which performs global optimization on the parameters of the optimized PID control algorithm and RBF neural network based on a genetic algorithm, and continuously adjusts the parameter combination through selection, crossover, and mutation operations to adapt to the operating requirements of the tensioning device under different working conditions; a data cache submodule, which adopts a double buffering mechanism to temporarily store the data before and after processing to ensure the continuity and stability of data processing and prevent data loss or processing interruption.

[0072] Specifically, the hierarchical parameter optimization processing center consists of a structure and functions of its various submodules. The data preprocessing submodule utilizes a sliding window-based median filter algorithm, setting appropriate parameters based on the data's noise characteristics to filter the raw multimodal data, effectively removing noise interference and improving data quality, laying the foundation for subsequent processing. The feature extraction submodule utilizes a method based on local mean decomposition to decompose multiple feature components from the preprocessed data, accurately extracting key characteristic parameters that reflect the tensioner's operating status, such as tension variation and displacement trends. The parameter optimization submodule, based on a genetic algorithm, uses selection, crossover, and mutation operations, mimicking biological evolution, to globally optimize the optimized PID control algorithm and RBF neural network parameters, continuously adjusting parameter combinations to adapt the algorithms to different operating conditions. The data caching submodule utilizes a double buffering mechanism to temporarily store both previous and subsequent data during processing, ensuring continuous and stable data processing and preventing data loss or processing interruptions. These submodules work together to achieve efficient processing of multimodal data and optimize algorithm parameters, improving the control accuracy and adaptability of the tensioner.

[0073] Preferably, the intelligent algorithm collaborative decision-making module includes a PID parameter adaptive adjustment unit, which monitors the operating status of the tensioning device in real time, and dynamically adjusts the proportional, integral, and differential parameters through the optimized PID control algorithm according to the collected data, so that the tensioning device responds quickly and reaches a stable operating state; an RBF neural network training unit, which uses historical operating data to perform offline training on the RBF neural network, and improves the prediction accuracy and generalization ability of the neural network by adjusting the center vector, width parameter and connection weight of the hidden layer node; a control strategy fusion unit, which performs weighted fusion on the control quantity output by the optimized PID control algorithm and the control quantity predicted by the RBF neural network, sets the weight coefficient according to different working conditions, and generates the final control strategy; a decision result output unit, which converts the generated control strategy into a standard control signal format, and transmits it to the adaptive variable frequency drive execution control unit through a data communication link.

[0074] Specifically, the intelligent algorithm collaborative decision-making module comprises the following components and functions: the PID parameter adaptive adjustment unit monitors the tensioner's operating status in real time and, based on collected data, dynamically adjusts the proportional, integral, and differential parameters using an optimized PID control algorithm. This ensures rapid response and stable operation in the face of changing operating conditions. The RBF neural network training unit utilizes extensive historical operating data to perform offline training of the RBF neural network. By adjusting the hidden layer node center vectors, width parameters, and connection weights, the neural network's prediction accuracy and generalization capabilities for different operating conditions are improved. The control strategy fusion unit combines the control variables output by the optimized PID control algorithm with those predicted by the RBF neural network, using weighted coefficients tailored to different operating conditions, to generate the final control strategy. The decision result output unit converts the generated control strategy into a standard control signal format and transmits it to the adaptive variable frequency drive execution and control unit via a data communication link. These units collaborate to achieve intelligent calculation and precise output of the tensioner's control strategy, ensuring stable and efficient operation under complex operating conditions.

[0075] Preferably, the adaptive variable frequency drive execution control unit includes a drive signal generation module, which generates a corresponding PWM drive signal according to the received control strategy to control the speed and torque of the drive motor; a power amplification module, which amplifies the generated PWM drive signal so that it has sufficient driving capability to drive the drive motor of the tensioning device to operate normally; a motor protection module, which monitors the current, voltage and temperature parameters of the drive motor in real time, and immediately cuts off the motor power supply when an abnormality is detected to prevent damage to the motor; a drive parameter configuration module, which allows the user to manually configure the rated speed, rated torque and acceleration and deceleration time parameters of the drive motor according to different application scenarios and load requirements of the tensioning device.

[0076] Specifically, the adaptive variable frequency drive execution and control unit comprises the following modules: the drive signal generation module receives the control strategy generated by the intelligent algorithm collaborative decision-making module and uses pulse width modulation technology to generate the corresponding PWM drive signal. This module precisely controls the drive motor speed and torque by adjusting the signal duty cycle. The power amplifier module utilizes high-power IGBT power devices to amplify the PWM drive signal, ensuring sufficient drive capability to ensure normal motor operation. The motor protection module monitors the drive motor's current, voltage, temperature, and other parameters in real time. Upon detecting an abnormality such as overcurrent, overvoltage, or overheating, the protection mechanism immediately triggers power cutoff to prevent motor damage. The drive parameter configuration module allows the user to manually configure the motor's rated speed, rated torque, acceleration and deceleration times, and other parameters based on the tensioner's different application scenarios and load requirements, enhancing the device's adaptability. These modules operate in tandem to achieve precise drive, effective protection, and flexible configuration of the drive motor, ensuring reliable operation of the tensioner.

[0077] like Figure 2 As shown, a control method for an intelligent variable frequency tensioning device includes the following steps:

[0078] Step S1: The multimodal data sensing and interaction integration unit collects the tension signal, displacement signal, and speed signal of the tensioning device in real time during operation according to a preset sampling frequency and communication protocol, and transmits the collected data to the hierarchical parameter optimization processing center via a high-speed serial bus;

[0079] Step S2: The hierarchical parameter optimization processing center performs data cleaning and feature extraction processing on the received data, uses a preset data processing algorithm to remove interference components in the data, extracts calibration feature parameters that can characterize the operating status of the tensioning device, and transmits the processed data to the intelligent algorithm collaborative decision module;

[0080] Step S3: The intelligent algorithm collaborative decision-making module calls the built-in optimized PID control algorithm and RBF neural network model based on the received processed data to calculate and make decisions on the control strategy of the tensioning device. In combination with various parameters of the tensioning device, the algorithm generates control instructions for the drive motor through iterative calculation;

[0081] Step S4: The control instruction generated by the intelligent algorithm collaborative decision module is transmitted to the adaptive variable frequency drive execution control unit, which converts the control instruction into a corresponding drive signal, and after power amplification processing, drives the drive motor of the tensioning device to operate, thereby driving the mechanical structure dynamic adjustment linkage unit to adjust the tensioning component;

[0082] Step S5: The state monitoring and feedback correction unit monitors the operating state of the tensioning device in real time, collects various state data during operation, and feeds the data back to the hierarchical parameter optimization processing center and the intelligent algorithm collaborative decision-making module for real-time correction of the control strategy and algorithm parameters;

[0083] Step S6: Repeat the above steps of data collection, processing, decision-making, execution and feedback correction to perform intelligent frequency conversion control and continuous stable operation of the tensioning device.

[0084] An intelligent variable frequency tensioning device and control method achieves comprehensive performance improvement through multi-module collaboration and deep integration of algorithms. At the control strategy level, the intelligent algorithm collaborative decision-making module combines the optimized PID control algorithm with the RBF neural network, completely changing the limitations of traditional fixed parameter PID control. The optimized PID control algorithm can dynamically adjust the proportional, integral, and differential parameters according to the real-time working conditions to achieve rapid response; the RBF neural network accurately predicts the control parameters under different working conditions by learning a large amount of historical data. The two complement each other, allowing the device to achieve high-precision tension and speed control under complex conditions such as sudden load changes and environmental interference, completely solving the problems of adjustment lag and large overshoot of traditional control strategies.

[0085] In terms of the coordination between mechanical structure and system, the multimodal data perception and interaction integration unit collects multi-dimensional data such as tension, displacement, and speed in real time. The hierarchical parameter optimization processing center performs in-depth feature extraction and preliminary processing on the data. The status monitoring and feedback correction unit monitors the operating status of the device in real time and feeds back data. The adaptive variable frequency drive execution control unit works closely with the mechanical structure dynamic adjustment linkage unit to accurately drive the motor and adjust the tensioning components according to the control instructions to ensure efficient linkage between the various components. Compared with the defects of traditional tensioning devices with poor coordination between the mechanical structure and drive system and imperfect monitoring and feedback mechanisms, the various modules of this device achieve fast and stable data interaction through specific data transmission protocols and communication links, and can obtain operating status information in real time, providing a reliable basis for the optimization and adjustment of the control strategy, greatly improving the overall operating efficiency and stability.

[0086] Furthermore, this intelligent variable-frequency tensioning device also boasts significant advantages in its detailed design. The data preprocessing submodule within the hierarchical parameter optimization processing hub utilizes a sliding window median filter algorithm to remove noise, the feature extraction submodule employs local mean decomposition to extract key features, the parameter optimization submodule employs a genetic algorithm for global optimization, and the data cache submodule employs a double buffering mechanism to ensure data continuity and stability. The adaptive variable-frequency drive execution and control unit's drive signal generation, power amplification, motor protection, and drive parameter configuration modules each perform their respective functions, comprehensively guaranteeing the device's stable operation. These designs give the device greater environmental adaptability and compatibility with operating conditions, meeting modern industry's demand for high-precision, intelligent tensioning devices.

[0087] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "disposed," "installed," "connected," "connected," and "fixed" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; they may refer to mechanical connections or electrical connections; they may refer to direct connections or indirect connections through an intermediate medium; and they may refer to internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0088] While embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent variable frequency tensioning device, characterized in that: include: Multimodal data perception and interaction integration unit, which establishes a two-way data transmission link with external tension sensors, displacement sensors, and speed sensors through a high-speed serial bus, and is used to collect real-time tension signals, displacement signals, and speed signals during the operation of the tensioning device; an adaptive variable frequency drive execution control unit, which is connected to the multimodal data perception and interaction integration unit through the industrial Ethernet protocol, receives the data signal transmitted by the perception unit, and performs variable frequency speed control on the drive motor of the tensioning device based on the preset variable frequency drive parameters; a hierarchical parameter optimization processing center, whose input port is connected to the data output end of the multimodal data perception and interaction integration unit, and whose output port is respectively connected to the data input end of the adaptive variable frequency drive execution control unit and the intelligent algorithm collaborative decision-making module, performs feature extraction and preliminary processing on the collected multimodal data; an intelligent The intelligent algorithm collaborative decision-making module establishes data communication channels with the hierarchical parameter optimization processing center and the adaptive variable frequency drive execution and control unit respectively, and has a built-in optimized PID control algorithm and RBF neural network model, and calculates and makes decisions on the control strategy based on the processed data; the mechanical structure dynamic adjustment linkage unit, whose control signal input end is connected to the control signal output end of the adaptive variable frequency drive execution and control unit, is connected to the tensioning component of the tensioning device through a transmission mechanism, and dynamically adjusts the tensioning component according to the received control signal; the state monitoring and feedback correction unit establishes data interaction links with the multimodal data perception and interaction integration unit, the hierarchical parameter optimization processing center, and the intelligent algorithm collaborative decision-making module respectively, monitors the operating status of the tensioning device in real time, and feeds back the monitoring data to the corresponding module for parameter correction and control strategy adjustment; The tension control model constructed based on the optimized PID control algorithm and RBF neural network is: in, The target tension value output by the tensioning device; is the proportional coefficient, which is set according to the load characteristics and mechanical structure parameters of the tensioning device; is the deviation between the actual tension value and the target tension value at the current moment; is the integral time constant, which is determined by the response speed requirement of the tensioning device; is the differential time constant, which is adjusted based on the dynamic characteristics of the tensioner; is the number of nodes in the hidden layer of the RBF neural network, which is set according to the control accuracy requirements of the tensioning device; The hidden layer The connection weight from each node to the output layer; is the radial basis function, is the input vector, including real-time collected tension, displacement, and velocity data. For the The center vector of hidden layer nodes, is the deviation between the actual tension value and the target tension value; In the intelligent algorithm collaborative decision-making module, the speed regulation model constructed based on the optimized PID control algorithm and RBF neural network is: in, The adjustment speed value of the tensioning device drive motor; is the proportional coefficient, which is determined according to the power parameters of the drive motor and the transmission ratio of the tensioning device; is the deviation between the actual speed value and the target speed value at the current moment; is the integral time constant, which is set according to the starting characteristics of the drive motor; is the differential time constant, which is adjusted by the speed regulation accuracy requirement of the drive motor; is the number of nodes in the hidden layer of the RBF neural network, which is set based on the dynamic response requirements of speed control; The hidden layer The connection weight from each node to the output layer; is the radial basis function, is the input vector, including the speed and load torque data collected in real time, For the The center vector of hidden layer nodes, It is the deviation between the actual speed value and the target speed value.

2. The intelligent variable frequency tensioning device according to claim 1, characterized in that: In the hierarchical parameter optimization processing center, the model for feature extraction of multimodal data is: in, is the extracted feature data vector; is the original tension data, A feature extraction function for tension data is constructed based on the sampling frequency and accuracy parameters of the tension sensor of the tensioning device; is the original displacement data, To extract the feature function of displacement data, it is set according to the travel range and resolution parameters of the tensioning device; is the original speed data, It is a feature extraction function for speed data, which is determined according to the speed range of the drive motor and the encoder accuracy parameters; Represents the fusion operation of feature data.

3. The intelligent variable frequency tensioning device according to claim 1, characterized in that: In the adaptive variable frequency drive execution control unit, the drive current regulation model constructed based on the optimized PID control algorithm and RBF neural network is: in, Adjust the current value for driving the motor; is the proportional coefficient, which is set according to the rated current and power factor parameters of the drive motor; is the deviation between the actual current value and the target current value at the current moment; is the integration time constant, which is determined by the overload capacity and heat dissipation characteristics of the drive motor; is the differential time constant, which is adjusted based on the response speed parameters of the drive circuit; is the number of hidden layer nodes of the RBF neural network, which is set according to the stability requirements of the drive current control; The hidden layer The connection weight from each node to the output layer; is the radial basis function, is the input vector, including the current, voltage, and load torque data collected in real time. For the The center vector of each hidden layer node.

4. The intelligent variable frequency tensioning device according to claim 1, characterized in that: In the state monitoring and feedback correction unit, the abnormal state warning model constructed based on the optimized PID control algorithm and RBF neural network is: in, It is an abnormal status flag. When the value is 1, it indicates an abnormality, and when the value is 0, it indicates normal. To monitor the real-time tension data, It is an abnormality judgment function based on tension data, which is constructed according to the rated tension and safety factor parameters of the tensioning device; is the threshold of abnormal tension; To monitor the real-time displacement data, It is an abnormality judgment function based on displacement data and is set according to the limit stroke parameters of the tensioning device; is the displacement anomaly threshold; To monitor real-time speed data, It is an abnormality judgment function based on speed data, which is determined according to the rated speed and speed regulation range parameters of the drive motor; is the speed abnormality threshold.

5. The intelligent variable frequency tensioning device according to claim 1, characterized in that: The hierarchical parameter optimization processing center includes a data preprocessing submodule, which performs filtering on the collected original multimodal data and adopts a median filtering algorithm based on a sliding window to remove noise interference in the data; The feature extraction submodule uses a feature extraction method based on local mean decomposition to decompose the multimodal data into multiple feature components and extract calibration feature parameters that can reflect the operating status of the tensioning device; The parameter optimization submodule performs global optimization on the parameters of the optimized PID control algorithm and RBF neural network based on the genetic algorithm. Through selection, crossover, and mutation operations, it continuously adjusts the parameter combination to adapt to the operating requirements of the tensioning device under different working conditions. The data cache submodule adopts a double buffering mechanism to temporarily store the data before and after processing to ensure the continuity and stability of data processing and prevent data loss or processing interruption.

6. The intelligent variable frequency tensioning device according to claim 1, characterized in that: The intelligent algorithm collaborative decision-making module includes a PID parameter adaptive adjustment unit, which monitors the operating status of the tensioner in real time and dynamically adjusts the proportional, integral, and differential parameters through an optimized PID control algorithm based on the collected data, so that the tensioner responds quickly and reaches a stable operating state; The RBF neural network training unit uses historical operating data to perform offline training on the RBF neural network. By adjusting the center vector, width parameters, and connection weights of the hidden layer nodes, the prediction accuracy and generalization ability of the neural network are improved. The control strategy fusion unit performs a weighted fusion of the control quantity output by the optimized PID control algorithm and the control quantity predicted by the RBF neural network, setting weight coefficients according to different working conditions to generate the final control strategy. The decision result output unit converts the generated control strategy into a standard control signal format and transmits it to the adaptive variable frequency drive execution control unit through a data communication link.

7. The intelligent variable frequency tensioning device according to claim 1, characterized in that: The adaptive variable frequency drive execution control unit includes a drive signal generation module, which generates a corresponding PWM drive signal according to the received control strategy to control the speed and torque of the drive motor; The power amplifier module amplifies the generated PWM drive signal to drive the normal operation of the drive motor of the tensioning device; the motor protection module monitors the current, voltage, and temperature parameters of the drive motor in real time. When an abnormality is detected, it immediately cuts off the motor power supply to prevent motor damage; the drive parameter configuration module allows the user to manually configure the rated speed, rated torque, and acceleration and deceleration time parameters of the drive motor according to the different application scenarios and load requirements of the tensioning device.

8. A control method for an intelligent variable frequency tensioning device, characterized in that: The method is applied to an intelligent variable frequency tensioning device according to claim 1, comprising the following steps: Step S1: The multimodal data sensing and interaction integration unit collects the tension signal, displacement signal, and speed signal of the tensioning device in real time during operation according to a preset sampling frequency and communication protocol, and transmits the collected data to the hierarchical parameter optimization processing center via a high-speed serial bus; Step S2: The hierarchical parameter optimization processing center performs data cleaning and feature extraction processing on the received data, uses a preset data processing algorithm to remove interference components in the data, extracts calibration feature parameters that can characterize the operating status of the tensioning device, and transmits the processed data to the intelligent algorithm collaborative decision module; Step S3: The intelligent algorithm collaborative decision-making module calls the built-in optimized PID control algorithm and RBF neural network model based on the received processed data to calculate and make decisions on the control strategy of the tensioning device. In combination with various parameters of the tensioning device, the algorithm generates control instructions for the drive motor through iterative calculation; Step S4: The control instruction generated by the intelligent algorithm collaborative decision module is transmitted to the adaptive variable frequency drive execution control unit, which converts the control instruction into a corresponding drive signal, and after power amplification processing, drives the drive motor of the tensioning device to operate, thereby driving the mechanical structure dynamic adjustment linkage unit to adjust the tensioning component; Step S5: The state monitoring and feedback correction unit monitors the operating state of the tensioning device in real time, collects various state data during operation, and feeds the data back to the hierarchical parameter optimization processing center and the intelligent algorithm collaborative decision-making module for real-time correction of the control strategy and algorithm parameters; Step S6: Repeat the above steps of data collection, processing, decision-making, execution and feedback correction to perform intelligent frequency conversion control and continuous stable operation of the tensioning device.

Images