Planetary speed regulator adaptive control method and system based on Internet of Things

Abstract

The invention relates to the technical field of planetary speed governing machine control, and discloses a planetary speed governing machine self-adaptive control method and system based on the Internet of Things. The method comprises the steps that a planetary gear set state parameter set is obtained through a multi-path isolation acquisition circuit, a dynamic feature vector is obtained through FFT algorithm frequency domain solution, a distributed control instruction set is obtained through fusion processing of an Internet of Things cooperative control network, optimal control parameters are obtained through adjustment of a self-adaptive parameter optimization algorithm, and the optimal control parameters are obtained. And finally, self-adaptive control output is obtained through real-time regulation and control of a predictive maintenance decision module. According to the invention, the technical problem that adaptive control of the planetary speed regulator based on multi-dimensional state perception and cooperation of the Internet of Things cannot be realized in the prior art is solved, and the equipment operation efficiency and the fault early warning accuracy are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of planetary speed regulator control, and particularly relates to a planetary speed regulator adaptive control method and system based on Internet of Things. BACKGROUND

[0002] In the prior art, the planetary speed regulator is widely used in the industrial fields of wind power generation, cement production, steel smelting and the like as an important transmission equipment. Its control method is mainly based on the traditional PID control strategy and simple threshold monitoring technology. The traditional control system usually adopts a single-point sensor to monitor the input speed and output torque, adjusts the motor driving parameters through a preset control algorithm, and realizes the basic speed and torque control functions. In the aspect of fault detection, the prior art mainly relies on regular manual inspection and alarm mechanism based on fixed threshold. When the monitored parameters exceed the preset threshold, an alarm signal is triggered. Some advanced control systems begin to use frequency converters and servo drive technology to optimize the transmission performance by adjusting the motor voltage and frequency, and are equipped with basic data acquisition and display functions. In terms of networking, some modern planetary speed regulator control systems begin to integrate Ethernet interfaces and wireless communication modules, which can upload running data to a central monitoring system to realize remote monitoring and basic data recording functions.

[0003] However, the prior art has significant technical defects and application limitations. First, the traditional control method lacks deep perception ability of the internal running state of the planetary gear set, and cannot accurately monitor the load distribution and meshing state among the sun gear, planetary gear and inner gear ring, resulting in insufficient control accuracy and delayed fault warning. Second, the existing single-machine control mode cannot realize the collaborative optimization among multiple planetary speed regulators, and is difficult to adapt to load changes and working condition fluctuations in complex industrial environments, resulting in low overall system efficiency. Third, the traditional regular maintenance and passive fault handling method cannot perform predictive maintenance according to the actual running state of the equipment, often resulting in over-maintenance or insufficient maintenance, increasing the operation and maintenance cost and affecting the production continuity. Fourth, the existing control algorithm mostly uses fixed parameter settings, lacks self-learning and adaptive ability, and cannot dynamically adjust the control strategy according to the equipment aging degree and environmental changes, resulting in gradually reduced transmission efficiency in long-term operation. Finally, the existing data processing method is mainly limited to simple statistical analysis and threshold comparison, lacks deep mining and intelligent analysis ability of multi-dimensional data, and cannot fully utilize the advantages of Internet of Things technology to realize intelligent management and optimized control of the equipment. SUMMARY

[0004] The present application provides a planetary speed regulator adaptive control method and system based on Internet of Things, which solves the technical problem that the prior art cannot realize adaptive control of the planetary speed regulator based on multi-dimensional state perception and Internet of Things collaboration, and improves the equipment running efficiency and fault warning accuracy.

[0005] In a first aspect, the application provides a planet speed regulator adaptive control method based on the Internet of Things, which comprises: synchronously collecting and processing the sun gear rotation speed, the planetary gear load distribution and the inner tooth ring vibration frequency spectrum of the planet speed regulator through a multi-path isolation collection circuit to obtain a set of planetary gear set state parameters; performing frequency domain solving processing on the reduction ratio change amount and the backhaul gap measurement value according to the set of planetary gear set state parameters through an FFT algorithm to obtain a planet transmission dynamics characteristic vector; performing multi-node data fusion processing on the planet transmission dynamics characteristic vector through an Internet of Things collaborative control network to obtain a distributed control instruction set; performing dynamic adjustment processing on the motor drive parameters and the load matching coefficient according to the distributed control instruction set through an adaptive parameter optimization algorithm to obtain optimal control parameters of the planet speed regulator; and performing real-time regulation and control processing on the control actuators and the drive units through a predictive maintenance decision module to obtain planet speed regulator adaptive control output.

[0006] Optionally, the synchronously collecting and processing the sun gear rotation speed, the planetary gear load distribution and the inner tooth ring vibration frequency spectrum of the planet speed regulator through a multi-path isolation collection circuit to obtain a set of planetary gear set state parameters comprises: performing electromagnetic induction isolation processing on the rotation speed signal of the sun gear input shaft through a wideband current-voltage transformer to obtain an isolated rotation speed signal; inputting the isolated rotation speed signal into a low-noise operational amplifier for band-pass filtering processing to obtain pure sun gear rotation speed data; performing multi-point synchronous measurement processing on the load distribution of the planetary gear tray based on a pressure sensor array to obtain a planetary gear load distribution matrix; performing differential amplification processing on the vibration acceleration sensor signal of the inner tooth ring installation position to obtain inner tooth ring three-axis vibration frequency spectrum data; performing timestamp synchronous merging processing on the pure sun gear rotation speed data, the planetary gear load distribution matrix and the inner tooth ring three-axis vibration frequency spectrum data to obtain the set of planetary gear set state parameters.

[0007] Optionally, the performing frequency domain solving processing on the reduction ratio change amount and the backhaul gap measurement value according to the set of planetary gear set state parameters through an FFT algorithm to obtain a planet transmission dynamics characteristic vector comprises: performing frequency domain decomposition processing on the set of planetary gear set state parameters through a fast Fourier transform algorithm to obtain a sun gear meshing frequency, a planetary gear passing frequency and an inner tooth ring resonance frequency; performing real-time calculation processing on the reduction ratio based on the sun gear meshing frequency and the planetary gear passing frequency to obtain a reduction ratio change amount; performing phase difference analysis processing on the inner tooth ring resonance frequency to obtain a planetary gear set back lash measurement value; performing time-frequency joint analysis processing on the reduction ratio change value and the back lash measurement value through a wavelet transform algorithm to obtain a planetary transmission frequency domain feature matrix; performing principal component analysis dimension reduction processing on the planetary transmission frequency domain feature matrix to obtain the planetary transmission dynamics feature vector.

[0008] Optionally, the planetary transmission dynamics feature vector is subjected to multi-node data fusion processing through an Internet of Things collaborative control network to obtain a distributed control instruction set, including: the planetary transmission dynamics feature vector is subjected to data encapsulation processing through an Ethernet controller to obtain a network transmission data packet; the network transmission data packet is subjected to multi-node broadcast transmission processing through a 4G wireless communication network to obtain distributed node receiving data; the distributed node receiving data is subjected to time synchronization correction and data integrity verification processing to obtain a synchronization feature vector set; the synchronization feature vector set is subjected to multi-source data integration processing through a weighted fusion algorithm to obtain a global optimization feature matrix; the global optimization feature matrix is subjected to control strategy generation processing through a distributed decision algorithm to obtain the distributed control instruction set.

[0009] Optionally, the motor drive parameters and load matching coefficients are dynamically adjusted through an adaptive parameter optimization algorithm according to the distributed control instruction set to obtain optimal control parameters of the planetary speed regulator, including: the distributed control instruction set is subjected to state space modeling processing through a model predictive control algorithm to obtain a planetary speed regulator dynamic state model; motor voltage, current and frequency parameters are subjected to Kalman filter estimation processing based on the planetary speed regulator dynamic state model to obtain an optimal motor drive parameter set; load torque and output power are subjected to matching degree calculation processing according to the optimal motor drive parameter set to obtain a load matching coefficient matrix; the load matching coefficient matrix is subjected to multi-objective function solving processing through a particle swarm optimization algorithm to obtain a control parameter optimization solution set; the control parameter optimization solution set is subjected to constraint condition verification and performance index evaluation processing to obtain the optimal control parameters of the planetary speed regulator.

[0010] Optionally, the load matching coefficient matrix is subjected to multi-objective function solving processing through a particle swarm optimization algorithm to obtain a control parameter optimization solution set, including: The load matching coefficient matrix is subjected to objective function construction processing to obtain a multi-objective function including maximization of transmission efficiency, minimization of energy consumption, and minimization of wear; Particle swarm initialization processing is performed based on the multi-objective function to obtain a particle swarm matrix including position vectors and velocity vectors; Adaptation value calculation processing is performed on each particle in the particle swarm matrix to obtain an individual optimal solution and a global optimal solution; Particle position and velocity are iteratively updated based on the individual optimal solution and the global optimal solution to obtain an optimized particle swarm; Convergence judgment and Pareto frontier extraction processing are performed based on the optimized particle swarm to obtain a control parameter optimization solution set.

[0011] Optionally, the planetary speed regulator optimal control parameters are subjected to real-time regulation and control processing on control actuators and drive units through a predictive maintenance decision module to obtain planetary speed regulator adaptive control output, including: The planetary speed regulator optimal control parameters are subjected to equipment health state prediction processing through a long short-term memory network algorithm to obtain a remaining useful life prediction value; Maintenance timing and failure risk are calculated based on the remaining useful life prediction value to obtain a predictive maintenance instruction; Control signals are generated based on the predictive maintenance instruction and the planetary speed regulator optimal control parameters to obtain actuator drive signals; The actuator drive signals are subjected to pulse width modulation and power amplification processing through a digital signal processor to obtain drive unit control voltages; The drive unit control voltages are used to perform real-time drive control processing on frequency converters and servo motors to obtain the planetary speed regulator adaptive control output.

[0012] In a second aspect, the application provides a planetary speed regulator adaptive control system based on the Internet of Things, including: The acquisition module is used to perform synchronous acquisition processing on the sun gear speed, planetary gear load distribution, and inner tooth ring vibration frequency spectrum of the planetary speed regulator through a multi-channel isolation acquisition circuit to obtain a planetary gear set state parameter set; The solving module is used to perform frequency domain solving processing on the reduction ratio change and the backstroke gap measurement value based on the planetary gear set state parameter set through an FFT algorithm to obtain a planetary transmission dynamics feature vector; The fusion module is used to perform multi-node data fusion processing on the planetary transmission dynamics feature vector through an Internet of Things collaborative control network to obtain a distributed control instruction set; An adjusting module is configured to perform dynamic adjusting processing on motor driving parameters and load matching coefficients by an adaptive parameter optimization algorithm according to the distributed control instruction set, so as to obtain optimal control parameters of the planetary speed regulating machine. A regulating module is configured to perform real-time regulating processing on control actuators and driving units by a predictive maintenance decision module, so as to obtain adaptive control outputs of the planetary speed regulating machine.

[0013] In a third aspect, a planetary speed regulating machine adaptive control device based on the Internet of Things is provided, which comprises a memory and at least one processor, and the memory stores instructions; the at least one processor invokes the instructions in the memory, so that the planetary speed regulating machine adaptive control device based on the Internet of Things executes the above-mentioned planetary speed regulating machine adaptive control method based on the Internet of Things.

[0014] In a fourth aspect, a computer readable storage medium is provided, which stores instructions, and when the instructions are executed on a computer, the computer executes the above-mentioned planetary speed regulating machine adaptive control method based on the Internet of Things.

[0015] In the technical scheme provided in the present application, the sun gear speed, the planetary gear load distribution and the inner tooth ring vibration frequency spectrum of the planetary speed regulating machine are synchronously collected and processed by the multi-path isolation collection circuit, which solves the technical problem that the prior art cannot deeply perceive the internal running state of the planetary gear set, and the purity and accuracy of signal collection are ensured through electromagnetic induction isolation of the wideband current-voltage transformer and band-pass filtering processing of the low-noise operational amplifier. The time-domain signal is converted into a frequency-domain feature through the frequency-domain solution processing of the reduction ratio change and the return gap measurement value by the FFT algorithm, and the sun gear meshing frequency, the planetary gear passing frequency and the inner tooth ring resonance frequency and other key parameters are accurately extracted, the dynamics characteristics of the planetary transmission system are effectively identified through the time-frequency joint analysis of the wavelet transform algorithm and the dimension reduction processing of the principal component analysis, and the feature extraction precision is higher and the noise suppression ability is stronger compared with the traditional time-domain analysis method. The limitation of traditional single machine control is broken through the multi-node data fusion processing of the Ethernet controller and the 4G wireless communication of the Internet of Things cooperative control network, the coordination and consistency between distributed nodes are ensured through time synchronization correction and data integrity verification, the global optimization control is realized through the weighted fusion algorithm and the distributed decision algorithm, and the overall efficiency of the cooperative operation of multiple planetary speed regulating machines is significantly improved. The adaptive parameter optimization algorithm combines model predictive control and Kalman filter estimation to dynamically adjust motor driving parameters and load matching coefficients, and finds the optimal balance point between maximum transmission efficiency, minimum energy consumption and minimum wear through the multi-objective function solution of the particle swarm optimization, which can automatically adjust the control strategy according to the actual working condition compared with the fixed parameter control mode, has stronger adaptability and higher control precision.

[0016] The predictive maintenance decision processing predicts the health state of the equipment through a long short-term memory network algorithm, uses the time sequence memory capability and nonlinear mapping characteristics of the LSTM network, accurately predicts the remaining useful life and generates maintenance instructions, avoids the blindness of traditional periodic maintenance and the hysteresis of passive fault processing, realizes accurate actuator control through pulse width modulation and power amplification of the digital signal processor, and finally forms the adaptive control output of the planetary speed regulator with the characteristics of fast response speed, high control precision and strong adaptability. In the field of wind power generation, the FFT algorithm can accurately identify the influence of wind load changes on gear meshing state, the Internet of Things collaborative control realizes the coordinated operation of multiple speed regulators in the wind farm, the particle swarm optimization algorithm optimizes the equipment life while ensuring the power generation efficiency, and the predictive maintenance of the LSTM network reduces the downtime loss caused by sudden failure. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor based on these drawings.

[0018] Figure 1 An embodiment schematic diagram of the adaptive control method of the planetary speed regulator based on the Internet of Things in the embodiments of the present application; Figure 2 An embodiment schematic diagram of the adaptive control system of the planetary speed regulator based on the Internet of Things in the embodiments of the present application; Figure 3 The structural schematic diagram of the adaptive control equipment of the planetary speed regulator based on the Internet of Things in the embodiments of the present application. DETAILED DESCRIPTION

[0019] The embodiments of the present application provide an adaptive control method and system of a planetary speed regulator based on the Internet of Things. The terms "first", "second", "third", "fourth" and the like (if any) in the specification and claims of the present application and the above drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the term "comprising" or "having" and any variation thereof is intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily limit to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0020] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the adaptive control method for planetary speed governors based on the Internet of Things in this application includes: Step S101: The sun gear speed, planetary gear load distribution and internal ring vibration spectrum of the planetary speed controller are synchronously acquired and processed through a multi-channel isolation acquisition circuit to obtain the planetary gear set state parameter set; Step S102: Based on the planetary gear set state parameter set, the reduction ratio change and backlash measurement values ​​are processed by the FFT algorithm in the frequency domain to obtain the planetary transmission dynamic feature vector. Step S103: The planetary transmission dynamics feature vector is processed by multi-node data fusion through an Internet of Things collaborative control network to obtain a distributed control instruction set; Step S104: Based on the distributed control instruction set, the motor drive parameters and load matching coefficients are dynamically adjusted using an adaptive parameter optimization algorithm to obtain the optimal control parameters for the planetary speed controller. Step S105: The optimal control parameters of the planetary speed governor are processed in real time by the predictive maintenance decision module to control the actuator and drive unit, thereby obtaining the adaptive control output of the planetary speed governor.

[0021] It is understood that the executing entity of this application can be an IoT-based planetary speed governor adaptive control system, or it can be a terminal or a server; the specific implementation is not limited here. This application's embodiments use a server as an example for illustration.

[0022] Specifically, the multi-channel isolation acquisition circuit uses a wideband current-voltage transformer to perform electromagnetic induction isolation processing on the rotational speed signal of the sun gear input shaft, converting the original voltage signal into an isolated rotational speed signal to avoid the influence of strong electrical interference on subsequent processing. The isolated rotational speed signal is then input into a low-noise operational amplifier for bandpass filtering to remove noise frequency components and retain the effective rotational speed change frequency range, resulting in clean sun gear rotational speed data. Simultaneously, a pressure sensor array performs multi-point synchronous measurement of the load distribution on the planetary gear tray, with each sensor corresponding to the load condition of one planetary gear, forming a load distribution matrix where each element represents the load value at the corresponding location. The vibration acceleration sensor signal at the internal gear ring mounting location is differentially amplified to eliminate common-mode interference, obtaining triaxial vibration spectrum data of the internal gear ring. These data are synchronously merged using timestamps to ensure consistency of all data in the time dimension, ultimately forming the planetary gear set state parameter set.

[0023] Fast Fourier Transform (FFT) converts the time-domain planetary gear set state parameters into a frequency-domain representation, extracting the sun gear meshing frequency, planetary gear pass-through frequency, and internal ring gear resonance frequency. Real-time reduction ratio calculation is determined by the ratio of the sun gear meshing frequency to the planetary gear pass-through frequency; as the sun gear speed changes, the corresponding frequency ratio changes, thus calculating the reduction ratio change. Phase difference analysis of the internal ring gear resonance frequency determines the phase delay caused by gear backlash by comparing the phase offset at each sampling point, thereby calculating the return clearance measurement. Wavelet transform algorithm performs time-frequency joint analysis on the reduction ratio change and return clearance measurement, decomposing the signal into different time and frequency scales to form a planetary transmission frequency domain feature matrix. Principal component analysis (PCA) reduces the dimensionality of this matrix, extracting the main feature components and removing redundant information to obtain the planetary transmission dynamics feature vector. The Ethernet controller encapsulates the planetary transmission dynamics feature vector according to the TCP / IP protocol, adding a header and checksum information to form a network transmission data packet. The 4G wireless communication network broadcasts the data packet to multiple control nodes, where each node receives the data and performs local storage and preliminary processing. Distributed nodes receiving data require time synchronization correction. GPS timing signals are used to unify the clocks of each node, and data integrity is verified to check for lost or corrupted data packets. The synchronized feature vector set is integrated using a weighted fusion algorithm. Data from different nodes is assigned different weight coefficients based on their reliability and accuracy, and a weighted average is calculated to form a globally optimized feature matrix. The distributed decision-making algorithm generates a control strategy based on this matrix, considering the control capabilities and load conditions of each node, allocating appropriate control tasks, and forming a distributed control instruction set.

[0024] Model predictive control (MMC) transforms a distributed control instruction set into a state-space model, establishing mathematical relationships between input, state, and output variables. Kalman filtering optimizes motor voltage, current, and frequency parameters, obtaining the optimal set of motor drive parameters through prediction and update phases, combining the system model and measurement data. Load matching is calculated based on the ratio of input to output power, combined with the torque characteristic curve, to determine the optimal matching coefficient under the current operating condition, forming a load matching coefficient matrix. Particle swarm optimization (PSO) constructs a multi-objective function encompassing transmission efficiency, energy consumption, and wear. Through iterative updates of the particle swarm's position and velocity, it searches for a Pareto optimal solution set. After constraint verification and performance evaluation, the optimal control parameters for the planetary speed governor are determined.

[0025] The Long Short-Term Memory (LSTM) network algorithm predicts the remaining service life of the equipment based on historical operating data and current control parameters. The network learns the equipment degradation patterns through information transmission between the input, hidden, and output layers. Predictive maintenance commands are generated based on the predicted remaining service life and fault risk assessment results. When the predicted service life falls below a set threshold, a maintenance command is triggered. Control signal generation combines the maintenance command with optimal control parameters to form a comprehensive actuator drive signal. The digital signal processor performs pulse width modulation on the drive signal, converting the digital control signal into an analog control voltage. A power amplifier circuit amplifies the signal strength to form the drive unit control voltage. Upon receiving the control voltage, the frequency converter and servo motor adjust their output frequency and torque to achieve precise control of the planetary speed controller, completing the entire adaptive control process and outputting the planetary speed controller adaptive control signal.

[0026] In one specific embodiment, the process of performing step S101 may specifically include the following steps: The rotational speed signal of the sun gear input shaft is electromagnetically isolated by using a wideband current and voltage transformer to obtain an isolated rotational speed signal. The isolated speed signal is input into a low-noise operational amplifier for bandpass filtering to obtain clean sun gear speed data. The load distribution of the planetary gear tray is measured synchronously at multiple points based on a pressure sensor array to obtain the planetary gear load distribution matrix. Differential amplification processing is performed on the vibration acceleration sensor signal at the mounting position of the internal gear ring to obtain the triaxial vibration spectrum data of the internal gear ring; The pure sun gear speed data, planetary gear load distribution matrix, and internal gear ring triaxial vibration spectrum data are time-stamped and merged to obtain the planetary gear set state parameter set.

[0027] Specifically, the wideband current-voltage transformer's electromagnetic induction isolation processing of the sun gear input shaft speed signal is based on the principle of electromagnetic induction. When the sun gear rotates, the voltage signal generated passes through the primary winding of the transformer, inducing a corresponding voltage signal in the secondary winding, achieving electrical isolation while maintaining the signal characteristics. The wideband characteristic ensures that the transformer can respond to a frequency range from DC to several megahertz, covering the signal frequencies generated by the sun gear at various speeds. The transformer's turns ratio is determined based on the input signal amplitude and the requirements of subsequent processing circuits, typically using a 1:1 ratio or a specific ratio design. After the isolated speed signal is processed by the transformer, power frequency interference and high-frequency noise in the original signal are effectively suppressed, while maintaining the integrity of the speed change information.

[0028] The low-noise operational amplifier employs an active filter design for bandpass filtering of the isolated speed signal. The input noise voltage of the operational amplifier is controlled at the microvolt level to ensure the amplification quality of weak signals. The passband frequency range of the bandpass filter is determined based on the operating speed range of the sun gear. The lower cutoff frequency is set to the fundamental frequency corresponding to the lowest speed, and the upper cutoff frequency is set to ten times the fundamental frequency of the highest speed, filtering out out-of-band noise. The filter gain is adjusted according to the input range of the subsequent ADC, amplifying the millivolt-level input signal to the volt-level output. The clean sun gear speed data retains the time-domain characteristics of speed changes and removes noise components introduced by power grid interference, electromagnetic radiation, and mechanical vibration.

[0029] Multi-point synchronous measurement of the load distribution on the planetary gear tray is achieved by mounting multiple high-precision pressure sensors on the tray. The number of sensors corresponds to the number of planetary gears, typically three or four. Each sensor measures the radial force transmitted to the tray by the corresponding planetary gear, and the voltage signal output by the sensor is proportional to the applied pressure. Synchronous measurement requires all sensors to acquire data at the same time, with the acquisition clock controlled by a unified time base signal to ensure time synchronization accuracy at the microsecond level. The load distribution matrix arranges the measured values ​​of each sensor according to their spatial position. The matrix elements reflect the load-bearing situation of each planetary gear; uneven load distribution will be reflected in the differences in element values ​​in the matrix.

[0030] The differential amplification processing of the vibration acceleration sensor at the internal gear ring mounting position employs a triaxial accelerometer to measure radial, tangential, and axial vibration accelerations respectively. The differential amplifier circuit converts the sensor's differential signal into a single-ended signal, eliminating the influence of common-mode noise. The amplification factor is set according to the vibration amplitude and measurement accuracy requirements. The triaxial vibration spectrum data is obtained by performing a frequency domain transformation on the time-domain vibration signal, extracting the amplitude and phase information of each frequency component. The natural frequency, meshing frequency, and harmonic components of the internal gear ring form characteristic peaks in the spectrum, and changes in the vibration spectrum reflect the working state and wear degree of the internal gear ring.

[0031] The timestamp synchronization and merging process aligns the clean sun gear speed data, planetary gear load distribution matrix, and internal ring gear triaxial vibration spectrum data according to their acquisition time. Timestamp accuracy is required to be at the millisecond level, using GPS timing or a high-precision crystal oscillator as the time reference. During data merging, different types of data are interpolated along a unified time axis to ensure complete state information is available at any given time. The planetary gear set state parameter set contains multi-dimensional state information, forming a data matrix describing the operating state of the planetary speed controller. Rows in the matrix represent time series, and columns represent different measurement parameters.

[0032] In one specific embodiment, the process of performing step S102 may specifically include the following steps: The state parameter set of the planetary gear set is decomposed in the frequency domain using the fast Fourier transform algorithm to obtain the meshing frequency of the sun gear, the passing frequency of the planetary gears, and the resonance frequency of the internal gear ring. The reduction ratio is calculated in real time based on the meshing frequency of the sun gear and the passing frequency of the planetary gear to obtain the change in the reduction ratio. Phase difference analysis was performed on the resonant frequency of the internal gear ring to obtain the measured value of the return clearance of the planetary gear set. The reduction ratio change and backlash measurement values ​​are processed by wavelet transform algorithm for time-frequency joint analysis to obtain the planetary transmission frequency domain feature matrix. Principal component analysis was performed on the frequency domain feature matrix of the planetary transmission to reduce its dimensionality, resulting in the planetary transmission dynamics feature vector.

[0033] Specifically, the Fast Fourier Transform (FFT) algorithm for the frequency domain decomposition of the planetary gear set's state parameter set converts the time-domain vibration, rotational speed, and load signals into frequency-domain representations. By performing a Discrete Fourier Transform (DFT) on the sampled data, the time-series data is decomposed into sine and cosine components of different frequencies. The algorithm first applies a window function to the input time-domain data to reduce spectral leakage, then performs a radix-2 FFT to convert the N-point time-domain sequence into an N-point frequency-domain sequence, where each frequency point represents the amplitude and phase information at a specific frequency. The sun gear meshing frequency corresponds to the periodic impact frequency generated when the sun gear meshes with the planetary gears; its value is equal to the sun gear rotational speed multiplied by the number of teeth on the sun gear, and it appears as a distinct spectral peak in the spectrum. The planetary gear passing frequency is the frequency generated when the planetary gears revolve around the sun gear; it is equal to the sun gear rotational speed divided by the number of planetary gears, reflecting the periodicity of the planetary gear set's motion. The internal gear ring resonance frequency is the inherent vibration frequency generated by the internal gear ring structure under external excitation; it is usually located in the higher frequency band and forms a resonance peak in the spectrum.

[0034] The real-time reduction ratio calculation is based on the mathematical relationship between the sun gear meshing frequency and the planetary gear passing frequency. The actual reduction ratio is calculated from the frequency ratio. The theoretical reduction ratio is equal to the ratio of the number of teeth on the internal ring gear to the number of teeth on the sun gear plus one, while the actual reduction ratio is calculated from the ratio of the meshing frequency to the passing frequency. The change in reduction ratio is determined by the difference between the actual and theoretical reduction ratios, reflecting the impact of gear wear, increased backlash, or assembly errors on the transmission ratio. When gear wear occurs, the effective number of teeth changes, causing a shift in the meshing frequency, which in turn affects the accuracy of the reduction ratio. The accuracy of frequency measurement directly affects the accuracy of the reduction ratio calculation; high-resolution spectrum analysis technology is used to ensure that the frequency identification accuracy reaches two decimal places.

[0035] Phase difference analysis of the resonant frequency of the internal gear ring determines the influence of gear backlash on vibration propagation by comparing the phase relationships of vibration signals at different measuring points. Phase difference analysis compares the phases of sensors at different positions at the same frequency, calculates the phase delay time, and then calculates the gap distance based on the propagation speed of the vibration wave in the gear structure. The return gap measurement is obtained by multiplying the phase delay time by the vibration propagation speed, with units of degrees in minutes or millimeters. The presence of gear backlash causes phase lag in the vibration signal during propagation, and the degree of lag is proportional to the gap size. Multi-point phase measurement involves arranging multiple sensors along the circumference of the internal gear ring to obtain phase information at different positions, and the spatial distribution of the phase difference determines the uniformity of the gap.

[0036] The wavelet transform algorithm employs continuous wavelet transform to decompose the signal into different time and frequency scales for the joint time-frequency analysis of the reduction ratio change and backlash measurement. The algorithm selects a suitable mother wavelet function and performs multi-scale analysis of the signal through scaling and translation operations to obtain the time-frequency distribution map. The time-frequency characteristics of the reduction ratio change reflect the temporal evolution of gear wear, with low-frequency components corresponding to long-term trends and high-frequency components corresponding to short-term fluctuations. The time-frequency analysis of the backlash measurement reveals the periodic characteristics of the backlash change and its correlation with load cycles and temperature variations. The joint time-frequency analysis compares the two parameters in the same time-frequency coordinate system to identify their correlation and causal relationships. The planetary transmission frequency domain feature matrix arranges the time-frequency analysis results according to the frequency and time dimensions, with matrix element values ​​representing the energy density at the corresponding time and frequency points.

[0037] Principal component analysis (PCA) reduces the dimensionality of the planetary transmission frequency domain feature matrix by calculating the matrix's covariance matrix and eigenvalue decomposition to extract the main feature directions. The algorithm first centers the matrix by subtracting the mean vector, then calculates the covariance matrix and solves for the eigenvalues ​​and eigenvectors. The magnitude of the eigenvalues ​​reflects the variance contribution of the data along the corresponding eigenvector directions; the eigenvectors corresponding to the largest eigenvalues ​​are selected as principal components. Dimensionality reduction projects the original high-dimensional feature matrix into a low-dimensional principal component space, preserving the main variation information while removing redundant and noise components. The planetary transmission dynamics eigenvectors contain the most important dynamic feature information; their dimension is much smaller than the original feature matrix, but they retain most of the effective information.

[0038] In one specific embodiment, the process of executing step S103 may specifically include the following steps: The planetary drive dynamics feature vectors are encapsulated and processed by an Ethernet controller to obtain network transmission data packets. Based on the network transmission of data packets, multi-node broadcast transmission processing is performed through a 4G wireless communication network to obtain distributed node received data; Time synchronization correction and data integrity verification are performed on the data received by the distributed nodes to obtain a set of synchronized feature vectors. The synchronized feature vector set is integrated from multiple sources using a weighted fusion algorithm to obtain a globally optimized feature matrix. Based on the globally optimized feature matrix, a distributed decision-making algorithm is used to generate control strategies, resulting in a distributed control instruction set.

[0039] Specifically, the Ethernet controller encapsulates the planetary drive dynamics feature vector data based on the TCP / IP protocol stack, layering the feature vector data according to network transmission protocol requirements. The physical layer handles electrical signal transmission; the data link layer adds Ethernet frame headers and trailers containing source and destination MAC addresses; the network layer adds IP headers containing source and destination IP addresses; and the transport layer adds TCP or UDP headers containing port numbers and sequence numbers. The feature vector data, as the application layer payload, is fragmented during encapsulation according to the Maximum Transmission Unit (MTU) limit, with each data packet's size controlled to within 1500 bytes to ensure reliable network transmission. Data encapsulation also includes checksum calculations, using CRC checksums to detect data errors during transmission. The network transmission data packet contains complete protocol header information and feature vector payload data.

[0040] 4G wireless communication networks employ the LTE protocol standard for multi-node broadcast transmission of data packets, establishing wireless connections between base stations and multiple control nodes. The broadcast transmission mode simultaneously sends the same data packet to all target nodes in the network, avoiding the duplicate transmission problem inherent in point-to-point transmission. The physical layer of the 4G network uses OFDMA multiple access technology, allocating spectrum resources to different user equipment. Data packets are converted into radio frequency signals by a modem at the air interface for transmission. The network protocol stack includes RRC connection management, PDCP data compression, RLC retransmission control, and MAC scheduling management, ensuring reliable transmission of data packets in the wireless channel. Distributed node data reception consists of broadcast data packets received by each planetary speed controller control node through its 4G module. Each node identifies its own data content based on its IP address and port number.

[0041] The time synchronization correction process for distributed node data reception is based on Network Time Protocol (NTP) or Precision Time Protocol (PTP), synchronizing with a GPS time server or network time server. Time synchronization correction calculates the deviation between each node's received time and the standard time, adjusting the data timestamps using linear or quadratic interpolation methods to ensure data alignment across all nodes on the timeline. Data integrity verification compares the hash value of the received data with the hash value of the sending end to check for data loss or errors during transmission. Lost data packets are retransmitted, erroneous data packets are corrected using error correction codes, and uncorrectable packets are marked as invalid. The synchronized feature vector set contains valid feature vector data from all nodes under a unified time base, forming a global data view of the distributed system.

[0042] The weighted fusion algorithm integrates multi-source data into a synchronized feature vector set, assigning weight coefficients based on the reliability and accuracy of each node's data. Weight calculation considers factors such as the historical data quality of nodes, sensor accuracy, network latency, and signal strength, calculating global feature values ​​using a weighted average method. During data fusion, feature vector components of the same dimension are weighted and summed, with the sum of the weight coefficients equal to 1, ensuring a reasonable numerical range for the fusion result. Anomaly detection uses statistical analysis to identify data points deviating from the normal range, employing the 3σ criterion or quartile method to remove outliers and avoid their impact on the fusion result. The globally optimized feature matrix integrates feature information from all valid nodes. Rows correspond to different feature dimensions, columns to different time points, and each matrix element represents the optimized value of that feature at a specific time within the global scope.

[0043] The distributed decision-making algorithm generates control strategies based on a globally optimized feature matrix and employs a multi-agent collaborative decision-making method. The algorithm first analyzes patterns and trends in the feature matrix to identify the planetary speed governor groups requiring coordinated control. Then, it formulates personalized control strategies based on the current state and load distribution of each governor. A decision tree or rule base contains control rules for different operating conditions, and the algorithm finds the most suitable combination of control rules for the current state through pattern matching. Control strategy optimization considers multiple objectives, including maximizing overall efficiency, load balancing, and extending equipment lifespan, and solves for the optimal combination of control parameters using a multi-objective optimization algorithm. The distributed control instruction set contains specific control commands sent to each control node. Each command includes the target node identifier, control parameter values, and execution time, and is sent back to each planetary speed governor control unit via a reverse data transmission path.

[0044] In one specific embodiment, the process of executing step S104 may specifically include the following steps: The distributed control instruction set is processed into a state-space model using a model predictive control algorithm to obtain the dynamic state model of the planetary speed governor. Based on the dynamic state model of the planetary speed controller, Kalman filtering is used to estimate the motor voltage, current and frequency parameters to obtain the optimal set of motor drive parameters; The load torque and output power are matched based on the optimal motor drive parameter set to obtain the load matching coefficient matrix. The load matching coefficient matrix is ​​processed by multi-objective function solution using particle swarm optimization algorithm to obtain the optimized solution set of control parameters; The optimal control parameters of the planetary speed governor are obtained by verifying the constraints and evaluating the performance of the optimized solution set of control parameters.

[0045] Specifically, the model predictive control algorithm (MMCC) performs state-space modeling of the distributed control instruction set by establishing a mathematical model of the planetary speed governor to describe its dynamic characteristics. The state-space model represents the operating state of the planetary speed governor as a state vector, containing key parameters such as the sun gear speed, planetary gear load, internal gear ring vibration amplitude, and reduction ratio. The input vector includes motor voltage, current, and frequency control signals, and the output vector represents the actual speed and torque response. The modeling process first establishes a system of differential equations based on the mechanical structural characteristics of the planetary gears to describe the mechanical and kinematic relationships between the components. Then, the continuous-time model is discretized into a discrete-time model suitable for digital controller processing. The state equations describe how the current state evolves to the next state based on the input signal, while the observation equations describe how the observable output signal is calculated from the internal state. The dynamic state model of the planetary speed governor determines the model parameters through system identification methods and trains the model using historical operating data to minimize the error between the model output and the actual measured values.

[0046] Kalman filtering estimates motor voltage, current, and frequency parameters based on optimal estimation theory, obtaining the optimal estimate by fusing system model predictions and sensor measurements. The Kalman filtering algorithm comprises two main stages: prediction and update. In the prediction stage, the state value and covariance matrix are predicted based on the state-space model and the previous optimal estimate. In the update stage, the predicted value and actual measured value are weighted and fused to obtain the current optimal estimate. The weighting coefficients are calculated based on the prediction error covariance matrix and the measurement noise covariance matrix; when the uncertainty of the predicted value is large, the measured value has a higher weight, and vice versa. Motor voltage parameter estimation considers the effects of power supply voltage fluctuations and load changes. Current parameter estimation fuses current sensor measurements and calculated values ​​based on the load model. Frequency parameter estimation combines the inverter's output frequency setpoint and the feedback value of the actual motor operating frequency. The optimal motor drive parameter set contains the filtered optimal estimates of voltage, current, and frequency, removing the effects of measurement noise and model uncertainty.

[0047] The load matching degree calculation process analyzes the degree of matching between the motor output characteristics and load requirements based on the optimal motor drive parameter set. The calculation process first calculates the actual output power based on the motor's voltage and current parameters, then calculates the load power requirement based on the speed and torque parameters, and evaluates the matching degree through the power ratio. The load torque calculation is based on the transmission characteristics of the planetary gear set and the actual load conditions at the output end, considering the influence of gear transmission efficiency and mechanical losses. The output power calculation combines the motor's efficiency characteristic curve and the current operating point position; the efficiency value varies under different speed and load conditions. The matching degree evaluation also includes the analysis of dynamic response characteristics, comparing the matching relationship between the motor's acceleration capability and the load change rate, as well as the matching relationship between the motor's speed regulation range and the load adjustment requirements. The load matching coefficient matrix arranges the matching degree values ​​under different operating conditions according to two dimensions: speed and load. The closer the matrix element value is to 1, the better the matching degree; the degree of deviation from 1 reflects the severity of poor matching.

[0048] The Particle Swarm Optimization (PSO) algorithm employs a swarm intelligence approach to find the optimal combination of control parameters for solving the multi-objective function of the load matching coefficient matrix. The algorithm first establishes a multi-objective function based on three objectives: maximizing transmission efficiency, minimizing energy consumption, and minimizing wear. Then, it initializes a particle swarm, with each particle representing a set of control parameter values. The particle's position vector corresponds to the numerical value of the control parameters, and its velocity vector corresponds to the trend of parameter changes. The fitness function calculates the performance index for each particle based on the multi-objective function. During optimization, each particle updates its velocity and position based on its historical best position and the swarm's global best position. The velocity update includes three components: inertia, cognition, and society, corresponding to the particle's motion inertia, individual learning ability, and swarm cooperation ability, respectively. The iterative process continues until the convergence condition is met or the maximum number of iterations is reached. Convergence is determined based on the magnitude of change of the optimal solution over several consecutive generations. The optimized control parameter solution set contains multiple Pareto optimal solutions, each representing the optimal parameter combination under different objective weights, forming a candidate set of control parameter schemes.

[0049] Constraint verification involves a feasibility check on each candidate solution in the optimized control parameter solution set to ensure that parameter values ​​are within the safe operating range of the equipment. Constraints include hard constraints such as motor voltage not exceeding 110% of rated voltage, current not exceeding 120% of rated current, frequency within the inverter's output range, speed not exceeding the gear's critical speed, and torque not exceeding the bearing's load capacity. Soft constraints include performance requirements such as transmission efficiency not less than 90% of the design value, temperature rise not exceeding insulation class limits, and vibration intensity within permissible ranges. Performance evaluation uses a comprehensive scoring method to rank candidate solutions that meet the constraints. Scoring indicators include control accuracy, response speed, stability, and robustness. Weighted scoring assigns weight coefficients to each indicator according to its importance, calculates the overall score, and selects the solution with the highest score as the optimal control parameters. The optimal control parameters for the planetary speed controller include the motor drive's voltage, current, and frequency setpoints, as well as the controller's gain parameters. These parameters have undergone multi-objective optimization and constraint verification, satisfying both performance requirements and ensuring equipment safety.

[0050] In one specific embodiment, the process of performing multi-objective function solving of the load matching coefficient matrix using a particle swarm optimization algorithm can specifically include the following steps: By processing the load matching coefficient matrix to construct the objective function, a multi-objective function is obtained that includes maximizing transmission efficiency, minimizing energy consumption, and minimizing wear. The particle swarm initialization process is performed based on a multi-objective function to obtain a particle swarm matrix containing position and velocity vectors; The fitness values ​​of each particle in the particle swarm matrix are calculated to obtain the individual optimal solution and the global optimal solution. The particle positions and velocities are iteratively updated based on the individual optimal solutions and the global optimal solutions to obtain the optimized particle swarm. Based on the optimized particle swarm, convergence judgment and Pareto front extraction are performed to obtain the optimized solution set of control parameters.

[0051] Specifically, the objective function construction process transforms the load matching coefficient matrix into an objective function expression for a mathematical optimization problem. Three optimization objectives are established by analyzing the transmission performance indicators reflected in the matrix. The transmission efficiency maximization objective function is based on efficiency-related elements in the load matching coefficient matrix, calculating the ratio of output power to input power, considering factors such as gear meshing losses, bearing friction losses, and lubricating oil viscosity losses. A higher objective function value indicates higher transmission efficiency. The energy consumption minimization objective function is established based on motor power consumption and auxiliary equipment energy consumption, including motor body losses, inverter losses, and cooling device energy consumption. Actual energy consumption is calculated by multiplying voltage, current, and power factor. A smaller objective function value indicates lower overall energy consumption. The wear minimization objective function is established based on gear contact stress, sliding speed, and lubrication conditions. Gear surface wear is calculated using the Arcard wear model, considering the impact of uneven load distribution and dynamic load impacts on accelerated wear. A smaller objective function value indicates longer equipment lifespan. The multi-objective function combines the three single-objective functions with weighted coefficients determined according to actual application requirements, forming a comprehensive performance evaluation index.

[0052] The particle swarm initialization process generates an initial particle swarm based on the variable dimensions and value ranges of the multi-objective function. Each particle represents a possible solution in the control parameter space. The position vector contains control variables such as motor voltage setpoint, current limit, frequency adjustment range, and PID controller gain parameters. Each component of the vector corresponds to a specific control parameter value. The position vector initialization uses a uniformly distributed random sampling method to ensure that particles are evenly distributed throughout the search space, avoiding the initial solution from being concentrated in a local region. The velocity vector represents the particle's movement direction and step size in the search space. The initial velocity is generated by random numbers. The magnitude of the velocity controls the particle's search range; too high a velocity leads to divergence in the search process, while too low a velocity results in slow convergence. The particle swarm matrix arranges all particle position and velocity vectors in rows. The number of rows in the matrix equals the number of particles, and the number of columns equals the dimension of the control parameters. The matrix element values ​​are randomly distributed within a preset parameter value range. The swarm size is determined based on the problem complexity and computational resources. Too few particles affect the global search capability, while too many increase the computational burden.

[0053] The fitness value calculation process substitutes the position vector of each particle in the particle swarm matrix into a multi-objective function for performance evaluation. The calculation process first decodes the particle position vector into specific control parameter values, then inputs these parameters into the simulation model or actual equipment of the planetary speed governor to calculate the corresponding transmission efficiency, energy consumption, and wear indicators. The multi-objective fitness evaluation uses a dominance relationship judgment method: a solution is said to dominate another solution when it is not inferior to another solution in all objectives and is superior to another solution in at least one objective. Individual optimal solutions record the best positions encountered by each particle during the historical search process. The current fitness value is compared with the historical best value to determine whether to update the individual optimal solution. The global optimal solution is selected from the individual optimal solutions of all particles. For multi-objective optimization problems, the global optimal solution is a set of solutions rather than a single solution, containing all currently found non-dominated solutions. The fitness value calculation also needs to consider the degree of constraint violation; solutions that do not meet the constraints are penalized by lowering their fitness value.

[0054] Iterative update processing adjusts the position and velocity vectors of particles based on individual optimal solutions and global optimal solutions, simulating the particle's motion in the search space. The velocity update formula consists of three parts: an inertia term that maintains the particle's original motion trend, a cognitive term that moves the particle towards its own historical optimal position, and a social term that moves the particle towards the global optimal position. The inertia weight controls the degree to which the particle maintains its current velocity; a larger weight makes the particle more likely to maintain its current search direction, while a smaller weight makes the particle more likely to change its search direction. The learning factor controls the degree to which the particle learns from the optimal solution; the cognitive learning factor affects the particle's movement towards its individual optimal solution, and the social learning factor affects the particle's movement towards the global optimal solution. Position update adds the current position to the updated velocity to obtain a new position vector. During the update process, it is necessary to check whether the position exceeds the value range; components exceeding the range need to be handled for boundary conditions. After optimization, the particle swarm includes all particles that have undergone one iteration update, and the distribution of particles gradually concentrates towards the optimal solution region, gradually improving the fitness level of the swarm.

[0055] Convergence assessment and Pareto front extraction evaluate the termination conditions and solution set quality of the optimization algorithm. Convergence assessment is based on the magnitude of change in the optimal solution over several consecutive generations; the algorithm is considered convergent when the magnitude of change is less than a preset threshold. The maximum number of iterations also limits the algorithm's runtime. Convergence assessment also includes evaluating population diversity; when particles are overly concentrated in a certain region, they may get stuck in local optima, requiring mutation operations to increase population diversity. Pareto front extraction identifies all non-dominated solutions from the current particle population; these solutions constitute the Pareto optimal solution set for the multi-objective optimization problem. The front extraction process determines non-dominated solutions by comparing the dominance relationships of all solutions pairwise. There are no dominance relationships between non-dominated solutions, and each solution is optimal on at least one objective. The distribution evaluation of the solution set measures the uniformity of solution distribution by calculating the distance between adjacent solutions. A uniformly distributed solution set provides decision-makers with more choices. The control parameter optimization solution set contains all combinations of control parameters corresponding to the Pareto optimal solutions. Each solution represents a different trade-off between the three objectives of transmission efficiency, energy consumption, and wear.

[0056] In one specific embodiment, the process of executing step S105 may specifically include the following steps: The optimal control parameters of the planetary speed governor are processed by a long short-term memory network algorithm to predict the health status of the equipment and obtain the predicted value of the remaining service life. Based on the predicted remaining useful life, decision-making calculations are performed on maintenance timing and failure risk to obtain predictive maintenance instructions; Based on the predictive maintenance instructions and the optimal control parameters of the planetary speed governor, control signals are generated and processed to obtain actuator drive signals; The actuator drive signal is processed by a digital signal processor to perform pulse width modulation and power amplification to obtain the drive unit control voltage; Based on the control voltage of the drive unit, the inverter and servo motor are driven and controlled in real time to obtain the adaptive control output of the planetary speed controller.

[0057] Specifically, the Long Short-Term Memory (LSTM) network algorithm for predicting the equipment health status of a planetary speed governor based on optimal control parameters utilizes a deep learning-based time series analysis method. It processes historical operating data by constructing a neural network structure containing input gates, forget gates, and output gates. The network's input layer receives time series data on the planetary speed governor's operating parameters, including the sun gear speed variation trend, historical data on planetary gear load distribution, the evolution of the internal gear ring vibration spectrum, and reduction ratio changes. The LSTM network's forget gate determines which information is discarded from the cell state. By analyzing the current input data and the previous hidden state, the activation value of the forget gate is calculated; an activation value close to 0 indicates complete forgetting, and close to 1 indicates complete retention. The input gate controls how new information is stored in the cell state, including the calculation of candidate values ​​and the input gate activation value. Candidate values ​​represent potential new information content, and the input gate activation value determines which parts of the candidate values ​​are added to the cell state. The output gate controls which parts of the cell state are output. The output gate activation value is calculated using the current input and the previous hidden state, and the activated cell state is then output as the current hidden state. The remaining useful life prediction is output through the fully connected layer, the last layer of the network. The output value represents the estimated remaining working time of the device in its current operating state. The prediction is based on a comprehensive analysis of historical degradation patterns and current state parameters.

[0058] The decision-making process calculates and processes the optimal maintenance timing based on the predicted remaining useful life, maintenance costs, and downtime losses. The calculation first establishes a mathematical model for maintenance decisions, considering factors such as preventative maintenance costs, fault repair costs, downtime losses, and spare parts inventory costs. An optimal maintenance threshold is determined through cost-benefit analysis. Failure risk assessment is based on the uncertainty of the predicted remaining useful life and statistical analysis of historical failure data. When the predicted useful life is below a safe threshold, the risk increases sharply; when the predicted useful life is sufficient, the risk remains at an acceptable level. The maintenance timing decision employs a dynamic programming method, comparing the total costs of immediate maintenance, delayed maintenance, and operation until failure, selecting the strategy with the lowest cost as the optimal decision. The decision calculation also considers constraints related to production plans and resource availability, avoiding maintenance operations during peak production periods and prioritizing maintenance during production downtime or planned downtime. Predictive maintenance instructions include information such as maintenance type, maintenance time, maintenance content, and resource requirements. The priority of instructions is determined based on the failure risk and its impact; maintenance instructions for high-risk equipment have higher execution priority.

[0059] The control signal generation and processing combines predictive maintenance commands with the optimal control parameters of the planetary speed governor to form a comprehensive control strategy. The generation process first analyzes the impact of maintenance commands on the current control parameters. When a maintenance command requires reduced load operation, the control parameters need to be adjusted accordingly to adapt to the load reduction requirement. When a maintenance command requires normal operation, the control parameters are executed according to the optimization results. Control signal generation considers the coordination of multiple control objectives, including the unification of performance, safety, and maintenance objectives, balancing conflicts between different objectives through a multi-objective decision-making method. Signal generation also includes the timing arrangement of control actions to ensure that the execution order of control commands conforms to the equipment operating procedures and avoids simultaneous execution of conflicting control actions. Actuator drive signals include motor control signals, brake control signals, lubrication device control signals, and monitoring device control signals, etc. Each signal has corresponding amplitude, frequency, and phase parameters, and the digital encoding of the signals adopts a standard industrial control protocol format.

[0060] The digital signal processor (DSP) converts the digital control signal into an analog power signal through pulse width modulation (PWM) and power amplification of the actuator drive signal. PWM technology controls the average power output by changing the pulse width. The selection of the modulation frequency needs to balance control accuracy and switching losses; too low a frequency leads to increased output ripple, while too high a frequency increases switching device losses. The duty cycle of the PWM signal is calculated based on the amplitude of the control command and is proportional to the output voltage. A digital comparator compares the reference signal with a triangular wave carrier to generate the PWM waveform. Power amplification uses IGBT or MOSFET power devices to amplify the low-power PWM signal into a high-power signal sufficient to drive the actuator. The amplifier design needs to consider conduction losses, switching losses, and heat dissipation requirements. The drive circuit includes an isolation driver, protection circuit, and feedback circuit. The isolation driver prevents interference between the main control circuit and the power circuit. The protection circuit automatically cuts off the output in case of overcurrent, overvoltage, and overheating. The feedback circuit monitors the deviation between the actual output and the command value. The control voltage of the drive unit is filtered to remove high-frequency switching noise. The accuracy and stability of the output voltage directly affect the control performance of the actuator.

[0061] Real-time drive control processing uses the drive unit control voltage to precisely control the frequency converter and servo motor. After receiving the control voltage, the frequency converter adjusts its output frequency and voltage amplitude, generating a three-phase AC voltage output through SPWM or SVPWM modulation technology. The frequency converter control employs vector control or direct torque control. Vector control decouples the torque and flux of the asynchronous motor, while direct torque control reduces control loops through direct control of flux linkage and torque. Servo motor control includes three control loops: position loop, speed loop, and current loop. The position loop controls the motor's angular position, the speed loop controls its speed, and the current loop controls its torque output. The control loop parameters are tuned using a classic PID controller or a state feedback controller based on modern control theory. PID parameters are determined through system identification and optimization algorithms, and the state feedback gain is designed using pole placement or optimal control theory. The planetary speed controller's adaptive control output is the final result of the entire control process, including the sun gear's speed control output, the planetary gear's load distribution output, the internal gear ring's vibration suppression output, and the overall transmission efficiency optimization output. The accuracy and response speed of the output signal directly determine the effectiveness of the adaptive control.

[0062] The above describes the IoT-based adaptive control method for planetary speed controllers in the embodiments of this application. The following describes the IoT-based adaptive control system for planetary speed controllers in the embodiments of this application. Please refer to [link to relevant documentation]. Figure 2 One embodiment of the IoT-based adaptive control system for planetary speed controllers in this application includes: The acquisition module is used to synchronously acquire and process the sun gear speed, planetary gear load distribution and internal ring vibration spectrum of the planetary speed controller through a multi-channel isolated acquisition circuit to obtain the planetary gear set state parameter set; The calculation module is used to perform frequency domain calculation on the reduction ratio change and backlash measurement values ​​based on the planetary gear set state parameter set using the FFT algorithm, so as to obtain the planetary transmission dynamic feature vector; The fusion module is used to perform multi-node data fusion processing on the planetary transmission dynamic feature vector through an Internet of Things collaborative control network to obtain a distributed control instruction set. The adjustment module is used to dynamically adjust the motor drive parameters and load matching coefficients according to the distributed control instruction set through an adaptive parameter optimization algorithm to obtain the optimal control parameters of the planetary speed controller. The control module is used to process the optimal control parameters of the planetary speed governor in real time through the predictive maintenance decision module to control the actuator and drive unit, thereby obtaining the adaptive control output of the planetary speed governor.

[0063] above Figure 2The adaptive control system for planetary speed controllers based on the Internet of Things (IoT) in this embodiment of the invention will be described in detail from the perspective of modular functional entities. The adaptive control device for planetary speed controllers based on the IoT in this embodiment of the invention will be described in detail from the perspective of hardware processing.

[0064] Reference Figure 3 This invention also provides an IoT-based adaptive control device for planetary speed controllers. This IoT-based adaptive control device can be a server, and its internal structure can be as follows: Figure 3 As shown, the IoT-based planetary speed governor adaptive control device includes a processor, memory, display screen, input device, network interface, and database connected via a system bus. The processor, designed as a computer, provides computing and control capabilities. The memory of the IoT-based planetary speed governor adaptive control device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the IoT-based planetary speed governor adaptive control device stores the data corresponding to this embodiment. The network interface of the IoT-based planetary speed governor adaptive control device is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements the above-described method.

[0065] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the IoT-based adaptive control device for planetary speed governors to which the present invention is applied.

[0066] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the Internet of Things-based planetary speed governor adaptive control method.

[0067] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0068] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an Internet of Things-based planetary speed controller adaptive control device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0069] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An adaptive control method for a planetary speed governor based on the Internet of Things, characterized in that, The method includes: The planetary gear set state parameter set is obtained by synchronously acquiring and processing the sun gear speed, planetary gear load distribution and internal ring vibration spectrum of the planetary speed controller through a multi-channel isolation acquisition circuit. Based on the state parameter set of the planetary gear set, the reduction ratio change and backlash measurement values ​​are processed in the frequency domain using the FFT algorithm to obtain the planetary transmission dynamic feature vector. The planetary transmission dynamics feature vectors are processed by multi-node data fusion through an Internet of Things collaborative control network to obtain a distributed control instruction set. Based on the distributed control instruction set, the motor drive parameters and load matching coefficients are dynamically adjusted using an adaptive parameter optimization algorithm to obtain the optimal control parameters for the planetary speed controller. The optimal control parameters of the planetary speed governor are processed in real time by the predictive maintenance decision module to regulate the control actuator and drive unit, thereby obtaining the adaptive control output of the planetary speed governor.

2. The adaptive control method for planetary speed governors based on the Internet of Things according to claim 1, characterized in that, The process involves synchronously acquiring and processing the sun gear speed, planetary gear load distribution, and internal ring gear vibration spectrum of the planetary speed controller via a multi-channel isolation acquisition circuit to obtain a set of planetary gear set state parameters, including: The rotational speed signal of the sun gear input shaft is electromagnetically isolated by using a wideband current and voltage transformer to obtain an isolated rotational speed signal. The isolated speed signal is input into a low-noise operational amplifier for bandpass filtering to obtain clean sun gear speed data. The load distribution of the planetary gear tray is measured synchronously at multiple points based on a pressure sensor array to obtain the planetary gear load distribution matrix. Differential amplification processing is performed on the vibration acceleration sensor signal at the mounting position of the internal gear ring to obtain the triaxial vibration spectrum data of the internal gear ring; The pure sun gear rotation speed data, the planetary gear load distribution matrix, and the internal gear ring triaxial vibration spectrum data are time-stamped and merged to obtain the planetary gear set state parameter set.

3. The adaptive control method for planetary speed governors based on the Internet of Things according to claim 1, characterized in that, The process involves using an FFT algorithm to perform frequency domain calculations on the reduction ratio change and backlash measurements based on the planetary gear set's state parameter set, resulting in a planetary transmission dynamics feature vector, including: The state parameter set of the planetary gear set is decomposed in the frequency domain using the fast Fourier transform algorithm to obtain the meshing frequency of the sun gear, the passing frequency of the planetary gears, and the resonance frequency of the internal gear ring. The reduction ratio is calculated in real time based on the meshing frequency of the sun gear and the passing frequency of the planetary gear to obtain the change in the reduction ratio. Phase difference analysis was performed on the resonant frequency of the internal gear ring to obtain the measured value of the return clearance of the planetary gear set. The reduction ratio change and the backlash measurement are processed by wavelet transform algorithm for time-frequency joint analysis to obtain the planetary transmission frequency domain feature matrix. Principal component analysis is performed on the frequency domain feature matrix of the planetary transmission to reduce its dimensionality, thereby obtaining the dynamic feature vector of the planetary transmission.

4. The adaptive control method for planetary speed governors based on the Internet of Things according to claim 1, characterized in that, The process of fusing multi-node data through an Internet of Things (IoT) collaborative control network to obtain a distributed control instruction set includes: The planetary transmission dynamics feature vector is encapsulated and processed by an Ethernet controller to obtain a network transmission data packet. Based on the data packets transmitted over the network, multi-node broadcast transmission processing is performed through the 4G wireless communication network to obtain distributed node received data. The data received by the distributed nodes is subjected to time synchronization correction and data integrity verification to obtain a set of synchronized feature vectors. The synchronized feature vector set is processed by multi-source data integration using a weighted fusion algorithm to obtain a globally optimized feature matrix. Based on the global optimized feature matrix, a distributed decision-making algorithm is used to generate control strategies, resulting in the distributed control instruction set.

5. The adaptive control method for planetary speed governors based on the Internet of Things according to claim 1, characterized in that, The process of dynamically adjusting the motor drive parameters and load matching coefficients based on the distributed control instruction set using an adaptive parameter optimization algorithm to obtain the optimal control parameters for the planetary speed controller includes: The distributed control instruction set is processed by state-space modeling using a model predictive control algorithm to obtain the dynamic state model of the planetary speed governor. Based on the dynamic state model of the planetary speed controller, Kalman filtering estimation is performed on the motor voltage, current and frequency parameters to obtain the optimal set of motor drive parameters; Based on the optimal motor drive parameter set, the load torque and output power are matched to obtain the load matching coefficient matrix. The load matching coefficient matrix is ​​processed by multi-objective function solution using particle swarm optimization algorithm to obtain the optimized solution set of control parameters; The optimal control parameters of the planetary speed governor are obtained by verifying the constraints and evaluating the performance of the optimized solution set of control parameters.

6. The adaptive control method for planetary speed governors based on the Internet of Things according to claim 5, characterized in that, The step of solving the load matching coefficient matrix using a particle swarm optimization algorithm to obtain a set of optimized control parameters includes: The load matching coefficient matrix is ​​processed to construct an objective function, resulting in a multi-objective function that includes maximizing transmission efficiency, minimizing energy consumption, and minimizing wear. Based on the multi-objective function, particle swarm initialization is performed to obtain a particle swarm matrix containing position vectors and velocity vectors; The fitness values ​​of each particle in the particle swarm matrix are calculated to obtain the individual optimal solution and the global optimal solution. The particle positions and velocities are iteratively updated based on the individual optimal solutions and the global optimal solutions to obtain an optimized particle swarm. Based on the optimized particle population, convergence judgment and Pareto front extraction are performed to obtain the optimized solution set of the control parameters.

7. The adaptive control method for planetary speed governors based on the Internet of Things according to claim 1, characterized in that, The step of using a predictive maintenance decision module to perform real-time adjustment and processing of the optimal control parameters of the planetary speed governor on the control actuator and drive unit to obtain the adaptive control output of the planetary speed governor includes: The optimal control parameters of the planetary speed governor are processed by a long short-term memory network algorithm to predict the health status of the equipment and obtain the predicted value of the remaining service life. Based on the predicted remaining useful life, decision-making calculations are performed on maintenance timing and failure risk to obtain predictive maintenance instructions; Based on the predictive maintenance instructions and the optimal control parameters of the planetary speed governor, control signal generation and processing are performed to obtain actuator drive signals; The actuator drive signal is processed by a digital signal processor for pulse width modulation and power amplification to obtain the drive unit control voltage. Based on the control voltage of the drive unit, the inverter and servo motor are driven and controlled in real time to obtain the adaptive control output of the planetary speed controller.

8. An adaptive control system for a planetary speed controller based on the Internet of Things, characterized in that, For implementing the IoT-based adaptive control method for planetary speed governors as described in any one of claims 1-7, the IoT-based adaptive control system for planetary speed governors includes: The acquisition module is used to synchronously acquire and process the sun gear speed, planetary gear load distribution and internal ring vibration spectrum of the planetary speed controller through a multi-channel isolated acquisition circuit to obtain the planetary gear set state parameter set; The calculation module is used to perform frequency domain calculation on the reduction ratio change and backlash measurement values ​​based on the planetary gear set state parameter set using the FFT algorithm, so as to obtain the planetary transmission dynamic feature vector; The fusion module is used to perform multi-node data fusion processing on the planetary transmission dynamic feature vector through an Internet of Things collaborative control network to obtain a distributed control instruction set. The adjustment module is used to dynamically adjust the motor drive parameters and load matching coefficients according to the distributed control instruction set through an adaptive parameter optimization algorithm to obtain the optimal control parameters of the planetary speed controller. The control module is used to process the optimal control parameters of the planetary speed governor in real time through the predictive maintenance decision module to control the actuator and drive unit, thereby obtaining the adaptive control output of the planetary speed governor.

9. An adaptive control device for a planetary speed controller based on the Internet of Things, characterized in that, It includes a memory and a processor, the memory storing a computer program that can run on the processor, and the processor executing the computer program to implement the Internet of Things-based adaptive control method for planetary speed governors as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is run by the processor, it causes the processor to execute the Internet of Things-based adaptive control method for planetary speed governors as described in any one of claims 1 to 7.

Images