Intelligent control method, device and equipment for street lamp cleaning process

Abstract

The invention discloses an intelligent control method, device and equipment for a street lamp cleaning process, and the method comprises the steps: collecting a street lamp surface pollution signal and an operation environment parameter, carrying out the reverse code conversion of the pollution signal through a self-adaptive sensing array, and generating a pollution driving response value; a vibration induction chain is constructed, multi-band vibration energy is obtained through resonance detection, and a power enhancement matrix is established; performing dead zone mining analysis on the matrix, determining a control gain position by adopting a dead zone wake-up technology, and forming an enhanced control node in combination with a pollution driving response value; self-adaptive parameters are obtained through adjustment of a fault presetting device, faults are converted into favorable control resources through defect compensation control, and a fault utilization chain is constructed; determining an operation period based on pollution inversion analysis, and executing pollution inversion operation to form a control enhancement field; and finally, gradient control processing is carried out through the response amplification window, an accurate cleaning control instruction is generated, intelligent and self-adaptive cleaning control of the street lamp equipment is achieved, and the cleaning efficiency and the energy utilization rate are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent cleaning control, in particular to a street lamp cleaning process intelligent control method, device and equipment. BACKGROUND

[0002] As an important part of urban infrastructure, street lamp equipment is exposed to complex outdoor environments for a long time. Its surface is easily eroded by various pollution sources such as atmospheric pollutants, dust, and bird droppings, resulting in a significant decrease in lighting efficiency and a reduction in service life, as well as an increase in maintenance costs and energy consumption. With the acceleration of urbanization and the intensification of environmental pollution, the pollution problem of street lamp equipment is becoming increasingly prominent, and traditional manual periodic cleaning methods cannot meet the fine requirements of modern urban management.

[0003] Traditional street lamp cleaning methods mainly rely on fixed-period manual work or simple timed cleaning systems, lacking real-time monitoring and intelligent analysis capabilities for pollution conditions. These methods cannot accurately identify the differences in pollution characteristics of different regions and different periods, making it difficult to achieve precise on-demand cleaning, resulting in serious waste of cleaning resources and uneven cleaning effects. Existing technologies generally have low pollution detection accuracy, single cleaning control strategies, and poor energy utilization efficiency. In particular, under complex environmental conditions, traditional methods lack intelligent control capabilities and cannot adaptively adjust to actual pollution conditions, making it difficult to achieve precise and efficient cleaning operations. SUMMARY

[0004] The present application discloses a street lamp cleaning process intelligent control method, device and equipment, aiming to solve the technical problems of low pollution detection accuracy, single cleaning control strategy, and poor energy utilization efficiency in traditional street lamp cleaning methods. By constructing a multi-source signal fusion pollution detection system, establishing an adaptive cleaning control strategy, and fully utilizing system fault features and environmental disturbance energy, the present application achieves intelligent and precise cleaning of street lamp equipment, improves cleaning efficiency, reduces energy consumption costs, and prolongs equipment service life.

[0005] The present application proposes a street lamp cleaning process intelligent control method in the first aspect, including the following steps: Collecting surface pollution signals and operating environment parameters of street lamp equipment, converting the surface pollution signals through an adaptive sensor array to obtain pollution gain data, extracting disturbance feature quantities from the operating environment parameters, and generating pollution driving response values by synergistically amplifying the pollution gain data and the disturbance feature quantities; Based on the pollution driving response values, a vibration induction chain is constructed, different frequency band vibrations are obtained by resonance detection of the vibration induction chain, energy collection points are extracted by vibration conversion of the different frequency band vibrations, and a dynamic enhancement matrix is constructed according to the energy collection points; The dead zone excavation analysis is performed on the power enhancement matrix to identify an activatable dead zone range, a dead zone wake-up determination control gain position is adopted from the activatable dead zone range, and the control gain position is coupled with the pollution driving response value to determine an enhancement control node; An adaptive parameter is determined based on the enhancement control node through fault presetter adjustment, a fault utilization chain is formed through defect compensation control by using the adaptive parameter, and a control amplification scheme is generated through synergistic fusion of the fault utilization chain; Pollution inversion analysis is performed on the surface pollution signal to obtain inversion depth data, an operation cycle is determined through the inversion depth data and the control amplification scheme, and a control enhancement field is formed based on the operation cycle through pollution inversion operation; A response amplification window is determined through the control enhancement field, and cleaning control instructions are obtained through control gradient processing based on the response amplification window.

[0006] The second aspect of the application proposes an intelligent control device for a street lamp cleaning process, comprising: A signal processing module is configured to collect surface pollution signals and operation environment parameters of a street lamp device, obtain pollution gain data through reverse coding conversion of the surface pollution signals by an adaptive sensing array, extract disturbance characteristic quantities from the operation environment parameters, and generate pollution driving response values through synergistic amplification of the pollution gain data and the disturbance characteristic quantities; A vibration processing module is configured to construct a vibration induction chain based on the pollution driving response values, obtain different frequency band vibrations through resonance detection of the vibration induction chain, extract an energy collection point set through vibration conversion of the different frequency band vibrations, and construct a power enhancement matrix according to the energy collection point set; A dead zone optimization module is configured to perform dead zone excavation analysis on the power enhancement matrix to identify an activatable dead zone range, determine a control gain position through dead zone wake-up from the activatable dead zone range, and determine an enhancement control node by coupling the control gain position with the pollution driving response value; An adaptive adjustment module is configured to determine an adaptive parameter based on the enhancement control node through fault presetter adjustment, form a fault utilization chain through defect compensation control by using the adaptive parameter, and generate a control amplification scheme through synergistic fusion of the fault utilization chain; An inversion control module is configured to perform pollution inversion analysis on the surface pollution signal to obtain inversion depth data, determine an operation cycle through the inversion depth data and the control amplification scheme, and form a control enhancement field based on the operation cycle through pollution inversion operation; An instruction generation module is configured to determine a response amplification window through the control enhancement field, and obtain cleaning control instructions through control gradient processing based on the response amplification window.

[0007] The third aspect of the present application provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the intelligent control method for a street lamp cleaning process disclosed in the first aspect when executing the program.

[0008] The beneficial effects of the present application are reflected in the following points: first, through the multi-source abnormal signal fusion technology, the surface pollution signal and the environmental parameters are analyzed cooperatively, and the pollution driving response value is generated by using the reverse coding conversion of the adaptive sensing array, which solves the problem of insufficient detection accuracy of traditional single sensor, realizes accurate quantification and real-time monitoring of pollution condition, so that the cleaning system can respond to pollution changes in time and make corresponding adjustment. Second, the vibration-induced chain and the power enhancement matrix are constructed, the originally invalid area in the system is converted into available control resources through the dead zone excavation and awakening technology, and the defect energy is converted into enhanced control power by combining the fault utilization chain, which avoids the waste of fault energy in the traditional method, realizes the full utilization of system resources and performance optimization. Finally, the pollution inversion analysis technology is used to construct the control enhancement field, and the accurate cleaning control instruction is generated through the gradient processing of the response amplification window, which changes the traditional fixed cycle cleaning mode, realizes the differentiated on-demand cleaning strategy, reduces unnecessary cleaning operation, reduces operation cost, and prolongs the maintenance interval and service life of the equipment through adaptive operation cycle adjustment.

[0009] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0010] The drawings herein show specific examples of the technical solutions described in the present application, and constitute part of the specification together with the specific embodiments, for explaining the technical solutions, principles and effects of the present application.

[0011] Unless specifically stated or defined otherwise, the same reference signs in different drawings represent the same or similar technical features, and different reference signs may also be used to represent the same or similar technical features.

[0012] Figure 1 is a flowchart of the intelligent control method for a street lamp cleaning process of the present application.

[0013] Figure 2 is a structural block diagram of the intelligent control device for a street lamp cleaning process of the present application.

[0014] Figure 3 is a structural diagram of the computer device of the present application. DETAILED DESCRIPTION

[0015] In the following description, for purposes of explanation and not limitation, specific details are set forth such as particular architectures, technologies, techniques, etc. in order to provide a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application can be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.

[0016] It is to be understood that the terminology "includes", "has", "holds", "contains" and / or "comprising", "including", "having" and their conjugates, as used herein, means "including but not limited to", and not to the exclusion of any other term or aspect.

[0017] It is also to be understood that the terminology "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items, as well as the lack of any associated listed items.

[0018] As used in the description of the application and the appended claims, the term "if' can be interpreted to mean "when" or "upon" or "in response to determining" or "in response to detecting" depending on the context. Similarly, the phrase "if it is determined" or "if [a described condition or event] is detected" can be interpreted to mean "upon determining" or "in response to determining" or "upon [the described condition or event] being detected" or "in response to [the described condition or event] being detected", depending on the context.

[0019] In addition, the terms "first", "second", "third", etc. as used in the description of embodiments herein and throughout the claims (if any), are not used to connote any relative importance but are merely to distinguish one element from another.

[0020] Reference throughout this specification to "one embodiment", "an embodiment", or "a specific embodiment", means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present application. Thus, appearances of the phrases "in one embodiment", "in an embodiment", "in some embodiments", "in other embodiments", "in additional embodiments", and so on, do not necessarily all refer to the same embodiment, unless otherwise indicated. The terms "including", "containing", "comprising", "having" and variations thereof, mean "including but not limited to", unless expressly specified otherwise.

[0021] The technical solutions of the embodiments of the present application are described below.

[0022] As Figure 1As shown, the embodiment of the present application provides a street lamp cleaning process intelligent control method, comprising the following steps S110-S160: Step S110, collecting the surface pollution signal and the working environment parameter of the street lamp equipment, obtaining the pollution gain data by reverse coding conversion of the surface pollution signal through the adaptive sensing array, extracting the disturbance characteristic quantity from the working environment parameter, and generating the pollution driving response value by cooperative amplification of the pollution gain data and the disturbance characteristic quantity.

[0023] Specifically, the surface pollution signal and the working environment parameter of the street lamp equipment are collected. By arranging multi-point contact pollution detection sensors on the surface of the street lamp pole, lampshade and radiator, the surface pollution conditions such as dust accumulation, oil adhesion and corrosion degree are monitored in real time. The pollution detection adopts the combination of capacitive sensor and infrared reflective sensor, the capacitive sensor detects the change of surface insulating layer to reflect the thickness of pollutants, and the infrared sensor detects the change of surface reflectivity to evaluate the properties of pollutants. The sensors are arranged in a cross configuration according to the annular array and the longitudinal array, the annular array is uniformly distributed around the circumference of the lamp pole, and the longitudinal array is arranged along the height direction of the lamp pole, forming a three-dimensional pollution monitoring network. The surface pollution signal contains multi-dimensional information such as pollutant thickness, distribution density, adhesion strength and chemical composition. The working environment parameters are collected synchronously, including environmental temperature, humidity, wind speed, wind direction, rainfall, atmospheric pressure and air quality index. The environmental parameters are obtained through a small weather station integrated on the top of the street lamp, and the weather station is equipped with temperature and humidity sensors, an anemorumbometer, a rain gauge and a particulate matter concentration detector. All sensor data are transmitted to the control center through the wireless communication module, and the control center is equipped with high-capacity storage devices and real-time data processing systems to ensure the synchronous collection and unified management of multi-source data.

[0024] The pollution gain data is obtained by reverse coding conversion of the surface pollution signal through the adaptive sensing array. The collected surface pollution signal is input into the adaptive sensing array processing system, which adopts a multi-channel parallel processing architecture, and each channel corresponds to the signal characteristics of one type of pollution. The reverse coding conversion adopts the back propagation algorithm to convert the pollution physical characteristic signal into standardized gain code. The mapping relationship between the pollution signal intensity and the gain coefficient is established as I=αS^β, wherein I is the pollution gain coefficient, S is the pollution signal intensity, α is the proportional coefficient, and β is the nonlinear index. The parameter values of different pollution types are determined through calibration experiments, the β value of dust pollution is about 1.2, the β value of oil pollution is about 1.5, and the β value of corrosion is about 1.8. The adaptive mechanism dynamically adjusts the coding parameters according to the environmental conditions, increases the humidity compensation factor in high humidity environment, and increases the thermal expansion compensation in high temperature environment. The coding conversion process adopts sliding average filtering to reduce noise interference, and the filter window length is adaptively adjusted according to the signal change rate.

[0025] The disturbance feature quantity is extracted from the working environment parameters. The collected multi-dimensional environment parameters are subjected to feature extraction and pattern recognition to identify the key environmental disturbance factors affecting the surface contamination of the street lamp. The comprehensive index THI = T + 0.36 x (RH-10) of temperature and humidity is calculated, where T is the temperature and RH is the relative humidity. The index reflects the comprehensive influence of temperature and humidity on the adhesion of pollutants. The time-varying characteristics of wind speed and direction are analyzed, and the wind field turbulence intensity and wind direction stability parameters are calculated, which affect the transmission and deposition process of pollutants. The instantaneous change rate and cumulative rainfall of rainfall intensity are extracted, which has a cleaning effect on the surface pollution but may also bring new pollutants. The gradient change of atmospheric particulate matter concentration is calculated to identify the environmental characteristics of pollution peak period and clean period. The disturbance feature quantity vector D = [THI, TI, WS, RF, PM] is established by comprehensively considering various environmental parameters, where TI is the turbulence intensity, WS is the wind direction stability, RF is the rainfall factor, and PM is the particulate matter factor.

[0026] In some embodiments, the pollution driving response value is generated by synergistically amplifying the pollution gain data and the disturbance feature quantity, including: extracting a pollution gain coefficient from the pollution gain data; constructing a gain-disturbance correlation curve using the disturbance feature quantity and the pollution gain coefficient; obtaining a stable amplification section in the gain-disturbance correlation curve; and determining the response value of the stable amplification section as the pollution driving response value.

[0027] The pollution gain coefficient is extracted from the pollution gain data. The obtained pollution gain data is subjected to numerical decomposition and feature parameter extraction to identify the effective gain component and noise component in the data. Digital filtering technology is used for pre-processing of the pollution gain data, and a Butterworth low-pass filter is used to remove high-frequency noise, with a filter cutoff frequency set to 1 / 10 of the sampling frequency. The local gain mean and variance are calculated by a moving window statistical analysis method, and the window size is adaptively adjusted according to the time scale of data variation. The maximum likelihood estimation method is used to extract the pollution gain coefficient, and a probability distribution model of the gain coefficient is established to solve the optimal gain coefficient by optimizing the likelihood function. The pollution gain data is classified according to the pollution type, and different types of pollution correspond to different gain coefficient extraction algorithms. The gain coefficient extraction of dust pollution focuses on cumulative effect analysis, the gain coefficient extraction of oil pollution focuses on adhesion strength analysis, and the gain coefficient extraction of corrosion focuses on chemical reaction rate analysis. The confidence interval of the gain coefficient is established to evaluate the reliability and accuracy of the extraction result. The extracted pollution gain coefficient is normalized to unify the gain coefficients of different dimensions into a standard interval, which facilitates subsequent correlation analysis and curve construction.

[0028] Exemplarily, the use of the disturbance characteristic quantity and the pollution gain coefficient to construct a gain-disturbance correlation curve includes: identifying disturbance fluctuation characteristics according to the disturbance characteristic quantity to determine a construction range, the disturbance fluctuation characteristics including disturbance amplitude, disturbance frequency and disturbance phase; tracking the gain change process along the construction range to form a gain trajectory based on the pollution gain coefficient; obtaining the characteristic coordinates of each correlation point in the gain trajectory; and constructing a gain-disturbance correlation curve by sorting the correlation strength of the characteristic coordinates.

[0029] The construction range is determined by identifying the disturbance fluctuation characteristics based on the disturbance signature. Frequency domain analysis is performed on the input disturbance signature. Fast Fourier transform is used to extract the signal's spectral characteristics and identify the dominant frequency components and harmonic components. The disturbance amplitude A = max(D) - min(D) is calculated, where D is the time series of the disturbance signature, reflecting the range of the disturbance intensity. The disturbance frequency f is determined through power spectral density analysis, and periodic disturbance patterns and random disturbance components are identified. The disturbance phase φ is extracted using the Hilbert transform, and the phase relationship and delay characteristics between different disturbance components are analyzed. A fluctuation characteristic parameter vector W = [A, f, φ] is established as the input parameter for determining the construction range. The vertical axis range of the correlation curve is determined based on the disturbance amplitude to ensure coverage of all possible gain coefficient values. The time window length for the correlation analysis is determined based on the disturbance frequency. High-frequency disturbances require shorter time windows, while low-frequency disturbances require longer time windows. The synchronization strategy for multivariate correlation analysis is determined based on the disturbance phase to compensate for the impact of phase delay on the correlation analysis. An adaptive range adjustment mechanism is established to automatically expand the construction range when the disturbance characteristics exceed the preset range, ensuring the integrity and representativeness of the correlation curve.

[0030] A gain trajectory is constructed by tracking the gain variation along the construction range based on the pollution gain coefficient. Within the defined construction range, disturbance signatures are sampled at equal or adaptive intervals, with each sampling point corresponding to a specific environmental disturbance state. For each sampled disturbance signature value, the corresponding gain response value is obtained from the pollution gain coefficient data through interpolation. Gain trajectory tracking employs Lagrange or spline interpolation to ensure trajectory continuity and smoothness. A step-size control algorithm is established for trajectory tracking, reducing the step size to improve accuracy in areas of drastic gain variation and increasing the step size to improve efficiency in areas of gentle variation. Key information from the trajectory tracking process is recorded, including local extreme values ​​of the gain coefficient, sudden rate of change points, and trend transition points. A mathematical model of the gain trajectory is fitted using multivariate regression analysis to establish a functional relationship between the gain coefficient and the disturbance signature. The gain trajectory is smoothed using a moving average filter or Gaussian filter to remove noise generated during tracking. The physical plausibility of the gain trajectory is verified, and its conformity to the basic laws and constraints of pollution physics is checked.

[0031] The feature coordinates of each associated point in the gain trajectory are obtained. The associated points with special significance on the constructed gain trajectory are identified, which reflect the key characteristics of the gain-perturbation relationship. The associated points include the starting point, the ending point, the extreme point, the inflection point, and the point with a significant change in slope. The first and second derivatives of each point on the trajectory are calculated using numerical differentiation methods, and the extreme points and inflection points are identified by the zero points of the derivatives. An importance evaluation index for associated points is established, considering the speciality of the point's location, the local slope characteristics, and the influence degree on the overall trajectory shape. The feature coordinates (Di, Gi) of each associated point are calculated, where Di is the perturbation characteristic value of the point, and Gi is the corresponding pollution gain coefficient value. The local feature parameters of the associated points are extracted, including local curvature, neighborhood slope, and influence radius, etc. A classification system of associated points is established, and the points are divided into different types such as enhancement points, suppression points, conversion points, and stable points, according to their characteristics. The distance and connection relationship between associated points are calculated, and the spatial distribution pattern and aggregation characteristics of the point group are analyzed. The importance of associated points is sorted, and the core associated points that contribute most to the gain-perturbation relationship are identified, providing key reference for the construction of associated curves.

[0032] The gain-perturbation association curve is constructed from the importance ranking of feature coordinates. Based on the extracted feature coordinates of associated points, the association strength of each point with the overall trend is calculated, and the contribution weight of each point to the final association curve is evaluated. The association strength calculation considers three dimensions: statistical significance, geometric importance, and physical reasonableness. Weighted least squares method is used for curve fitting of associated points, and the weight coefficient is determined according to the association strength. Points with high association strength obtain greater fitting weight. A piecewise fitting strategy is adopted, and the data is divided into several segments according to the distribution characteristics of associated points. Each segment uses a suitable fitting function. An optimization objective function for curve construction is established to find the optimal balance between fitting accuracy and curve smoothness. The generalization performance of different fitting schemes is evaluated through cross-validation method, and the curve model with the best performance on test data is selected. The final gain-perturbation association curve is constructed, and the curve equation form is G=Σwᵢfᵢ(D), where wᵢ is the weight coefficient, and fᵢ is the base function. The quality of the constructed association curve is evaluated, and indicators such as fitting goodness, residual distribution, and prediction accuracy are calculated. An uncertainty quantification system for the curve is established, and the reliability range of the curve is described through confidence interval and prediction interval.

[0033] Stable amplification segment in gain-disturbance correlation curve is obtained. The correlation curve is segmented and analyzed to identify different characteristic segments. The first derivative dG / dD and the second derivative d²G / dD² of the curve are calculated to analyze the slope change and concave-convex characteristics. The stable amplification segment is defined as the segment with positive first derivative and second derivative change amplitude less than the set threshold, which represents the interval where the pollution gain coefficient increases stably with the increase of the disturbance characteristic quantity. The sliding window method is used to detect the stability of the curve segment, and the standard deviation of the derivative value in the window is less than the stability threshold to determine the stable segment. The boundaries of the identified stable amplification segment are determined by finding the derivative mutation point to determine the start and end positions of the segment. The geometric characteristic parameters of the stable amplification segment are calculated, including the segment length, average slope, maximum slope and slope change rate. The stability evaluation index is established, which considers factors such as segment length, slope stability and data density. The multiple stable amplification segments are sorted and screened, and the most representative and reliable segment is selected as the basis for subsequent analysis.

[0034] The response value of the stable amplification segment is determined as the pollution driving response value. Based on the identified stable amplification segment, the product response value of the pollution gain coefficient and the disturbance characteristic quantity in the segment is calculated. The response value calculation uses the integral method R=∫[D1→D2]G(D)×DdD, where D1 and D2 are the boundary points of the stable amplification segment, and G(D) is the correlation curve function. The response value is calculated by numerical integration method, and the trapezoidal integration method or Simpson integration method is used to ensure the calculation accuracy. The calculated response value is dimensionless, and the response value of different physical quantities is converted into a standardized index. The response value is graded, and the pollution driving degree is divided into four grades: slight, moderate, severe and extremely severe. The time change rate of the response value is calculated to analyze the change trend and acceleration characteristics of the pollution driving intensity.

[0035] Step S120, based on the pollution driving response value, a vibration-induced chain is constructed, and the vibration-induced chain is resonant detected to obtain different frequency band vibrations, and the energy collection point set is extracted through vibration conversion of different frequency band vibrations, and a dynamic enhancement matrix is constructed according to the energy collection point set.

[0036] Specifically, a vibration induction chain is constructed based on the pollution-driven response value. The pollution-driven response value is input as a vibration excitation source, and a conversion relationship from the response value to vibration parameters is established. The pollution-driven response value is converted into vibration frequency and amplitude parameters through a nonlinear mapping function, and the mapping function adopts a power function form f=k×R^α, where f is the vibration frequency, R is the pollution-driven response value, k is the proportional coefficient, and α is the nonlinear index. According to the inherent frequency characteristics of the street lamp structure, a multi-stage vibration induction chain is established, the first stage is the base vibration at the bottom of the lamp pole, the second stage is the bending vibration at the middle of the lamp pole, and the third stage is the swing vibration of the lamp. Each stage of vibration forms a chain transmission relationship through structural coupling, and the lower level vibration excites the upper level vibration through a resonance amplification mechanism. The mathematical model of the induction chain is described by a multi-degree-of-freedom vibration system, and a mass-stiffness-damping matrix equation is established where [M], [C], and [K] are the mass, damping, and stiffness matrices, respectively, {x} is the displacement vector, and {F} is the excitation force vector. The modal frequency and mode shape of the vibration induction chain are determined by solving the eigenvalue problem, and the vibration mode that is easy to be excited is identified.

[0037] Resonance detection is performed on the vibration induction chain to obtain vibrations in different frequency bands. Sweep excitation tests are performed on the constructed vibration induction chain, and the excitation frequency is continuously changed from low frequency to high frequency, and the vibration response characteristics of each node are monitored. The resonance detection uses the frequency response function measurement method to calculate the transfer function H(ω)=X(ω) / F(ω) between the excitation point and the response point, where X(ω) is the frequency domain representation of the response, F(ω) is the frequency domain representation of the excitation, and ω is the angular frequency. When the transfer function has a peak value, it is determined as a resonance point, and the resonance frequency and resonance amplitude are recorded. The damping ratio of each order mode is calculated by the half-power point method, and the damping ratio ζ=Δω / (2ωr), where Δω is the half-power bandwidth, and ωr is the resonance frequency. The detected resonance frequencies are classified according to the frequency size, and the low frequency band (0-50Hz) mainly corresponds to the overall bending mode, the medium frequency band (50-200Hz) corresponds to the local bending and torsional mode, and the high frequency band (above 200Hz) corresponds to the high-order local mode. A frequency band classification database is established to record the characteristic parameters, excitation conditions, and response characteristics of each frequency band. The modal identification is performed on the vibrations in different frequency bands, and the modal parameter identification algorithm is used to extract the frequency, damping, and mode shape parameters of each order mode. The reliability of the identification result is evaluated by the modal confidence criterion to ensure the accuracy and integrity of the frequency band classification.

[0038] In some embodiments, the vibration conversion and energy extraction point set are extracted by the different frequency band vibrations, including: dividing the different frequency band vibrations into high frequency blocks and low frequency blocks; projecting an energy conversion channel from the high frequency blocks to the low frequency blocks; tracking the coordinates of power-rich points on the energy conversion channel to form a rich point group; and selecting the point with the strongest power in the rich point group to form an energy collection point set.

[0039] Vibrations in different frequency bands are divided into high-frequency and low-frequency blocks. Power spectral density analysis is used to calculate the energy contribution of each frequency component and identify frequency intervals with concentrated energy. A cutoff frequency f_c is set as the demarcation criterion between the high-frequency and low-frequency blocks. This cutoff frequency is determined by the energy equalization principle, ensuring that the total energy of the high-frequency and low-frequency blocks is essentially equal. The low-frequency block is defined as the vibration components in the frequency range 0 ≤ f ≤ f_c. This block contains the fundamental mode and main bending modes of the vibration-induced chain, with large vibration amplitudes and strong propagation capabilities. The high-frequency block is defined as the vibration components in the frequency range f > f_c. This block contains high-order local modes and harmonic components, with higher vibration frequencies and concentrated energy density. The vibration components within each block are subclassified: low-frequency blocks are divided into bending, torsional, and axial sub-blocks based on modal shape, while high-frequency blocks are divided into nodal, anti-nodal, and mixed sub-blocks based on local characteristics. A block characteristic parameter database is established to record the frequency range, energy distribution, dominant mode, and propagation characteristics of each block. The wavelet transform method is used to perform time-frequency analysis to identify the time-varying characteristics of high-frequency blocks and low-frequency blocks in non-steady-state vibration signals.

[0040] An energy conversion channel is projected from the high-frequency block to the low-frequency block. An energy transfer mechanism from the high-frequency block to the low-frequency block is established, and the topology and transfer path of the energy conversion channel are designed using nonlinear vibration theory. The energy conversion channel utilizes the principle of frequency down-conversion, converting high-frequency vibration energy into low-frequency vibration energy through nonlinear coupling elements. A mechanical frequency converter is designed, using a gear transmission or lever amplification mechanism to achieve frequency scaling. The conversion ratio r = f_high / f_low is determined based on specific application requirements. Nonlinear springs and dampers are installed at key nodes in the vibration induction chain to form a physical channel for energy conversion. The restoring force characteristic of the nonlinear spring is F = kx + k³x³, where k is the linear stiffness coefficient, k³ is the cubic nonlinear coefficient, and x is the displacement. The direction and efficiency of energy conversion are controlled by adjusting the nonlinear coefficient, achieving directional energy flow from the high-frequency block to the low-frequency block. A mathematical model of the energy conversion channel is established, and the energy transfer process within the channel is described using a set of coupled oscillator equations. Numerical simulation methods are used to analyze the channel's transfer characteristics and optimize the channel parameters for optimal energy conversion.

[0041] The coordinates of the power-rich points on the energy conversion channel are tracked to form a group of rich points. A power monitoring sensor array is deployed within the established energy conversion channel to measure the instantaneous power density and energy flow rate at each location in the channel in real time. A power-rich point is defined as a spatial location where the local power density exceeds twice the average value, and these points reflect the concentrated areas in the energy conversion process. The power distribution is measured point by point along the channel using a moving window scanning method, and the window size is determined according to the geometric dimensions of the channel and the vibration wavelength. A spatial distribution function P(x, y, z) of the power density is established, where x, y, and z are spatial coordinates, and P is the power density value. Through gradient analysis method Local maximum points of power density are identified, which correspond to candidate locations of power-rich points. A clustering analysis algorithm is used to group the identified rich points, and adjacent rich points with similar power levels are grouped into the same group. A feature description system of rich points is established, including spatial coordinates, power intensity, stability index, and influence range, etc.

[0042] The strongest points in the group of rich points are selected to form a set of energy harvesting points. The identified group of rich points is sorted by power intensity, and the power statistical characteristics of all rich points in each group are calculated. The group power evaluation uses a weighted average method P_group = Σ(w_i × P_i), where w_i is the weight coefficient of the i-th rich point, and P_i is the corresponding power value. The weight is determined according to the stability and duration of the point. A grading standard of power intensity is established, and the rich points are divided into strong, medium, and weak levels according to the power size, and the strong level rich points are preferred as energy harvesting candidate points. A multi-criteria decision-making method is used to evaluate the comprehensive performance of the rich points, and the evaluation indexes include power intensity, location stability, installation feasibility, and maintenance convenience. An optimization model of the number of harvesting points is established to find the optimal balance point between harvesting effect and cost investment, and to determine the most suitable number of harvesting points. The spatial optimization layout of the selected energy harvesting points is performed to avoid mutual interference and energy competition between the harvesting points. A hierarchical management strategy of the harvesting points is designed, and the harvesting points are divided into main harvesting points, auxiliary harvesting points, and emergency harvesting points to establish a multi-level energy harvesting system.

[0043] A power enhancement matrix is constructed according to the set of energy harvesting points. Based on the determined set of energy harvesting points, the energy distribution and coordination relationship between each harvesting point is established, forming a systematic power enhancement strategy. The row dimension of the power enhancement matrix corresponds to different energy harvesting points, and the column dimension corresponds to different output power distribution schemes. The matrix element aij represents the power contribution weight of the ith harvesting point under the jth distribution scheme. The matrix constraint condition Σaij=1 is established through the energy balance equation to ensure that the total power distribution under each scheme is 100%. Genetic algorithm is used to optimize the matrix parameters, with the maximum total output power as the objective function, while constraining the system stability and reliability indicators. An adaptive adjustment mechanism for the matrix is established to dynamically optimize the weight distribution according to changes in environmental conditions and equipment status. A fault switching mechanism is established to automatically adjust the matrix parameters when a harvesting point fails, ensuring that the overall system performance is not severely affected.

[0044] In step S130, dead zone excavation analysis is performed on the power enhancement matrix to identify the range of activatable dead zones. From the range of activatable dead zones, a dead zone wake-up is used to determine the control gain position, and the control gain position is coupled with the pollution driving response value to determine the enhancement control node.

[0045] Specifically, dead zone excavation analysis is performed on the power enhancement matrix to identify the range of activatable dead zones. Based on the constructed power enhancement matrix, the areas in the matrix with energy distribution weights close to zero or insensitive responses are analyzed, which constitute the dead zone characteristics of the system. Dead zone excavation uses singular value decomposition method to decompose the power enhancement matrix A into where U and V are orthogonal matrices, and Σ is a diagonal matrix containing singular values. By analyzing the distribution characteristics of the singular values, the subspace corresponding to the singular values close to zero is identified, which reflects the low energy transmission area in the power enhancement matrix. A dead zone judgment criterion is established, which judges as a dead zone when the absolute value of the matrix element is less than a set threshold θ, which is determined according to the system noise level and control accuracy requirements. Connectivity analysis method is used to aggregate scattered dead zone elements into continuous dead zone range, and dilation and erosion algorithm in morphological operation is used to eliminate isolated points and fill small gaps. The range of activatable dead zones is defined as the dead zone area that can produce effective response under external excitation, and the activation potential of each dead zone is evaluated by applying a test excitation signal. An activation potential evaluation index P=ΔE / E_input is established, where ΔE is the energy increment after excitation, and E_input is the input excitation energy. The identified range of activatable dead zones is managed hierarchically, and is divided into easy activation zone, medium activation zone and difficult activation zone according to the activation difficulty and potential benefits.

[0046] In some embodiments, the determining the control gain position from the activatable dead zone range comprises: identifying a flow isolation terminal zone from the activatable dead zone range; calculating a gain potential from the flow isolation terminal zone; generating a continuous gain space from the gain potential; and determining the control gain position from the continuous gain space.

[0047] For example, the identifying a flow isolation terminal zone from the activatable dead zone range comprises: injecting a reverse excitation into the boundary of the activatable dead zone range to generate a response echo; analyzing the decay trajectory of the response echo to determine an impedance jump point; forming a flow isolation zone from the impedance jump point; and defining the flow isolation terminal zone from the terminal boundary of the flow isolation zone.

[0048] A reverse excitation is injected into the boundary of the activatable dead zone range to generate a response echo. A reverse excitation generating device is installed at the determined boundary position of the activatable dead zone range, which can generate an excitation signal in the opposite direction of the normal excitation. The reverse excitation adopts pulse modulation technology, and the excitation signal is a short pulse sequence. The pulse width and interval are determined according to the response time constant of the dead zone. The selection of the excitation injection point is based on the analysis results of the boundary impedance, and the boundary position with relatively low impedance and sensitive response is preferred. The strength of the reverse excitation adopts an adaptive adjustment strategy, starting from low strength and gradually increasing until a significant response echo signal is detected. The generation mechanism of the response echo is based on the reflection principle of waves. When the excitation wave encounters an impedance mismatch interface during propagation, it reflects to form an echo signal. A echo detection system is established, and a high-sensitivity sensor array is arranged near the excitation injection point to monitor the amplitude, phase and spectral characteristics of the echo signal in real time. Signal processing techniques are used to filter and amplify the echo signal to improve the signal-to-noise ratio and detection accuracy. The time-domain characteristics of the echo signal include arrival time, peak amplitude and decay rate, and the frequency-domain characteristics include main frequency component and harmonic distribution. A characteristic database of the echo signal is established to record the echo characteristics under different excitation conditions.

[0049] The attenuation trajectory of the response echo is analyzed to determine the impedance jump point. Detailed attenuation characteristic analysis is performed on the captured response echo signal, and the attenuation process of the echo on the propagation path is reconstructed by backtracking calculation. The attenuation trajectory analysis uses an exponential decay model A(t)=A0e^(-αt), where A(t) is the echo amplitude at time t, A0 is the initial amplitude, and α is the attenuation coefficient. The attenuation coefficient is determined by fitting the measured attenuation data, and the mutation position of the attenuation coefficient corresponds to the impedance jump point on the propagation path. Backtracking analysis uses time reversal technology to reverse the echo signal on the time axis and reconstruct the propagation process of the excitation wave inside the dead zone. An impedance distribution model Z(x)=Z0(1+β(x)) is established for the propagation path, where Z0 is the reference impedance and β(x) is the impedance variation function at position x. The impedance jump point is defined as the position where the impedance variation rate |dZ / dx| exceeds a certain threshold, which reflects the discontinuity of the structure inside the dead zone. A multi-resolution analysis method is used to analyze the attenuation trajectory at different time scales to identify fast-changing and slow-changing impedance characteristics. A classification system for impedance jump points is established, which is divided into strong jump points, weak jump points, positive jump points, and negative jump points according to the jump amplitude and direction. Statistical analysis is used to determine the spatial distribution and aggregation characteristics of the impedance jump points, and to identify the structure patterns inside the dead zone.

[0050] The impedance jump points are processed to form a flow blocking area. Based on the identified impedance jump point positions, a special response blocking mechanism is designed to physically or virtually prevent the transmission of response signals at this point. The response blocking process uses impedance matching principles to insert impedance adjustment elements at the jump point, making the impedance of the propagation path continuous, thereby reducing wave reflection and scattering. The parameter design of the blocking element uses transmission line theory, with an impedance value Z=√(L / C), where L is the inductance and C is the capacitance, and impedance matching is achieved by adjusting the values of L and C. Evaluation indicators for blocking effectiveness are established, including reflection coefficient, transmission coefficient, and insertion loss, to quantify the effectiveness of the blocking process. The flow blocking area is defined as the spatial region within the influence range of the blocking process, and the response flow in this area is significantly suppressed or redirected. Field analysis method is used to calculate the spatial distribution of the blocking area, and contour and three-dimensional distribution maps of the blocking intensity are established. The dynamic adjustment mechanism of the blocking process adjusts the blocking parameters in real time according to the changes in system working conditions, ensuring the stability and adaptability of the blocking effect. A multi-point collaborative blocking strategy is established, which is enabled when the single-point blocking effect is not good, and the blocking effect is superimposed through phase control.

[0051] The flow block termination zone is defined by the terminal boundary of the flow block region. Based on the flow block region, boundary analysis and terminal location determination are performed to identify the area range that responds to complete flow stop. The terminal boundary is determined using field strength analysis methods to calculate the response field strength distribution within the flow block region and identify the boundary location where the field strength drops to the background noise level. A boundary detection algorithm is established to identify the sharp change in response intensity by calculating the gradient ∇R, and these locations form the candidate points of the terminal boundary. Connectivity analysis methods are used to connect the scattered boundary points into a continuous boundary line, and spline curve fitting techniques are used to smooth the boundary. The definition of the flow block termination zone uses the closed region principle, and the closed region surrounded by the continuous terminal boundary is the flow block termination zone. The geometric feature description of the termination zone is established, including area, perimeter, centroid position, and principal axis direction parameters, providing a geometric basis for subsequent gain potential calculation. A multi-level analysis method is used to decompose the large termination zone into several sub-regions, each with relatively uniform characteristics and functions.

[0052] The gain potential calculation from the flow block termination zone constitutes the gain multiplier. In-depth analysis is performed within the identified flow block termination zone to evaluate the potential ability of the region to produce response gain under external excitation. The gain potential calculation uses the energy accumulation principle to analyze the potential energy stored in the termination zone and the releasable response capacity. An energy storage model E_stored = ∫Vε(r)dV is established, where ε(r) is the energy density at spatial location r, and V is the volume of the termination zone. By applying excitation signals of different intensities, the response characteristics of the termination zone are measured, and an excitation-response relationship curve is established. The gain multiplier is defined as the ratio of the response intensity after excitation R_after to the response intensity before excitation R_before, M = R_after / R_before, which quantifies the response amplification capacity of the termination zone. A multi-frequency scanning method is used to measure the gain characteristics of the termination zone at different frequencies to identify the frequency dependence and resonance characteristics of the response gain. A spatial distribution function M(x, y, z) of the gain multiplier is established to describe the differences in gain potential at different locations within the termination zone. Statistical analysis methods are used to calculate the mean, variance, and distribution characteristics of the gain multiplier to provide a data basis for subsequent response fitting.

[0053] The gain multiplier is obtained by response fitting to produce a continuous gain space. Based on the discrete data points of the calculated gain multiplier, mathematical fitting methods are used to construct a continuous gain distribution function, realizing the conversion from discrete points to continuous space. Response fitting uses polynomial fitting, spline interpolation, or radial basis function methods, and the most suitable fitting algorithm is selected according to the data characteristics. Fitting quality evaluation indicators are established, including fitting error, correlation coefficient, and residual analysis, to ensure the accuracy and reliability of the fitting results. The mathematical expression of the continuous gain space uses a piecewise function form G(x, y, z) = Σw i φi (x,y,z), where φ i are basis functions, w i are weight coefficients. The weight coefficients are determined by least squares or weighted least squares method to minimize the deviation of the fitting function from the measured data. The gradient field ∇G of the gain space is established to identify the direction and region with the most dramatic gain change, providing guidance information for gain location determination. A multi-resolution analysis method is used to hierarchically describe the gain space, revealing the detailed characteristics of the gain distribution at different scales.

[0054] The boundary extraction is performed in the continuous gain space to determine the control gain location. Contour extraction technology is used to draw the contours of different gain levels in the gain space, identifying the boundary region with the largest gain gradient. A gain threshold classification system is established to divide the continuous gain space into high-gain, medium-gain and low-gain regions, focusing on the boundary characteristics of the high-gain region. Morphological image processing methods are used to extract the boundary of the gain space, using edge detection operators to identify the location of gain jump changes. The selection of control gain location uses a multi-objective optimization strategy, considering factors such as gain intensity, location stability, implementation feasibility and control effect. A mathematical model of location optimization is established, with the objective function being to maximize the control gain while minimizing the number of locations and implementation costs. Heuristic search algorithms such as genetic algorithm, particle swarm algorithm or simulated annealing algorithm are used to solve the optimal location combination. Sensitivity analysis is performed on the determined control gain location to evaluate the degree of influence of small changes in location on control effect.

[0055] The control gain location is coupled with the pollution driving response value to determine the enhanced control node. The coupling model uses a nonlinear function form N=f(G,R)=αG^β+γR^δ+ηGR, where N is the intensity of the enhanced control node, G is the gain value of the control gain location, R is the pollution driving response value, and α, β, γ, δ, η are coupling parameters. The optimal values of the coupling parameters are determined by fitting experimental data, and the least squares method or genetic algorithm is used for parameter optimization. A quantitative indicator of coupling strength is established to evaluate the synergistic effect of control gain location and pollution driving response value. The spatial distribution of the enhanced control node is determined using a weighted fusion method, with the weight coefficients allocated according to the importance of the control gain location and the influence range of the pollution driving response value. An association network between nodes is established, and the connection relationship and information transmission path between nodes are analyzed by graph theory. The performance of the enhanced control node is evaluated, and an evaluation index system is established including response speed, control accuracy, stability and robustness. A hierarchical control strategy for nodes is designed, and nodes are divided into master nodes, auxiliary nodes and monitoring nodes according to their importance and functional characteristics. A fault diagnosis and fault tolerance mechanism for nodes is established, and when a node fails, a backup node is automatically enabled or the network topology is reconfigured.

[0056] Step S140 , adjusting the fault presetter based on the enhanced control node to determine adaptive parameters, using the adaptive parameters to form a fault utilization chain through defect compensation control, and performing collaborative integration on the fault utilization chain to generate a control amplification solution.

[0057] Specifically, adaptive parameters are determined by adjusting the fault presetter based on the enhanced control node. The fault presetter adopts a multi-mode switching architecture, including three operating states: prevention mode, monitoring mode, and compensation mode. It automatically switches operating modes based on the activation level of the enhanced control node. The presetter adjustment mechanism uses parameter adaptation technology to dynamically adjust system parameters by monitoring node response characteristics in real time. A mapping relationship θ=f(N,t) is established between node response and parameter adjustment, where θ is the adaptive parameter vector, N is the enhanced control node strength, and t is the time variable. Parameter adjustment uses a gradient descent method, with the adjustment step size adaptively determined based on the rate of change of the node response. Adaptive parameters include key control parameters such as the control gain coefficient, response time constant, damping coefficient, and nonlinear compensation coefficient. A parameter boundary constraint mechanism is established to ensure that the adjusted parameters are within the safe operating range, preventing system instability or performance degradation. A multivariable optimization method is used to simultaneously adjust multiple parameters, and the optimal solution set in the parameter space is found through Pareto optimization. A convergence criterion is designed for parameter adjustment, and the adjustment process is terminated when the parameter change rate falls below a set threshold.

[0058] In some embodiments, the use of the adaptive parameters to form a fault utilization chain through defect compensation control includes: establishing a parameter time series from the adaptive parameters; implementing defect marking on the parameter time series to form a compensation time node; using the compensation time node as a boundary to divide the parameter time series into a pre-compensation segment and a post-compensation segment; comparing the control characteristics of the pre-compensation segment and the post-compensation segment to form a fault utilization chain.

[0059] A parameter time series is established from the adaptive parameters. The sampling frequency of the parameter time series is determined based on the system's dynamic characteristics and control requirements to ensure that key characteristics and transient processes of parameter changes are captured. Multidimensional adaptive parameters are time-aligned to eliminate time offsets and sampling errors between different parameters and establish a unified time base. A sliding window technique is used to segment the parameter time series, with the window length determined based on the system's characteristic time constant and control period. A method for calculating the parameter change rate is established, and numerical differentiation techniques are used to calculate the parameter change rate and acceleration at each moment. Statistical analysis is performed on the parameter time series to calculate statistical characteristics such as mean, variance, skewness, and kurtosis, identifying the basic patterns of parameter distribution. Spectral analysis of the parameter time series is performed using frequency domain analysis to identify the periodic components and frequency characteristics of parameter changes. A noise filtering mechanism is established for the parameter time series, using digital filtering techniques to remove measurement noise and high-frequency interference.

[0060] The compensation time node is constituted by defect labeling on parameter time series. Defect labeling adopts multi-level detection strategy, including statistical anomaly detection, trend anomaly detection and pattern anomaly detection. Statistical anomaly detection judges by calculating the degree of parameter deviation from normal range, and marks as statistical anomaly when parameter value exceeds 3σ range. Trend anomaly detection identifies abnormal change mode by analyzing parameter change trend and slope, and marks sudden trend turning point or abnormal steep change as trend anomaly. Pattern anomaly detection adopts special identification method to train normal parameter mode identification system, and marks parameter change deviating from normal mode as pattern anomaly. A defect severity grading system is established, and defects are divided into four grades of slight, medium, severe and critical according to the influence degree on system performance. The compensation time node is defined as the time point when defect is detected and compensation control needs to be started, and the determination of the node considers the severity, development trend of defect and response ability of system. A time stamp recording system of node is established to accurately record the occurrence time, defect type and severity of each compensation time node.

[0061] The parameter time series is divided into pre-compensation segment and post-compensation segment by using compensation time node as boundary. The pre-compensation segment is defined as the parameter sequence between the last compensation end time and the current compensation time node, which reflects the natural evolution process of system without compensation intervention. The post-compensation segment is defined as the parameter sequence between the current compensation time node and the next compensation end time, which reflects the system response process under the action of compensation control. The length statistical analysis of time segment is established to calculate the average duration, maximum duration and time distribution characteristics of pre-compensation segment and post-compensation segment. The boundary effect processing technology is adopted to specially process the data near the segmentation boundary, avoiding the influence of segmentation operation on data continuity. The inter-segment correlation analysis is established to study the correlation between adjacent time segments by correlation analysis and causality analysis method. The parameter change in each time segment is modeled, and polynomial fitting or exponential function is used to describe the evolution law of parameters in the segment.

[0062] Comparing the control characteristics of the pre-compensation and post-compensation stages constitutes a fault utilization chain. An in-depth comparative analysis of the pre-compensation and post-compensation stages is conducted to quantify the improvement in system performance and the degree of fault utilization achieved through compensation control. This control characteristic comparison utilizes a multi-dimensional analysis approach, including key performance indicators such as response speed, control accuracy, stability, and energy efficiency. A performance improvement metric, PI = (P_after - P_before) / P_before, is established, where P_after represents the performance indicator of the post-compensation stage and P_before represents the performance indicator of the pre-compensation stage. Statistical analysis is used to calculate the magnitude of improvement and significance level for each performance indicator to confirm the effectiveness of compensation control. The effectiveness of fault utilization is analyzed using the energy conversion efficiency metric, calculating the conversion ratio of fault energy to useful control energy. A mathematical representation of the fault utilization chain is established to describe the causal relationship and transmission mechanism between fault detection, compensation control, and performance improvement. Time series analysis is used to examine the differences in the dynamic characteristics of the pre-compensation and post-compensation stages, identifying the impact of compensation control on the system's dynamic behavior.

[0063] The fault utilization chain is collaboratively integrated to generate a control amplification solution. The constructed fault utilization chain is collaboratively optimized with other control links of the system, and the overall amplification of the control effect is achieved through multi-link integration. The collaborative integration adopts a distributed control architecture, and each control link achieves global optimization through information exchange and coordination mechanisms. The objective function of the fusion optimization is established as J = Σw i f i (x i ), where w i is the weight coefficient, f i is the performance function of the ith link, x i is the corresponding control variable. Multi-agent coordination technology is used to solve the fusion optimization problem. Each control link acts as an agent, achieving optimal control through local information exchange and global goal coordination. The design of the control amplification scheme considers the multi-timescale characteristics of the system, adopting different amplification strategies at different time scales. An analysis system for amplification effects is established, including indicators such as improved response speed, improved control accuracy, increased stability margin, and enhanced robustness.

[0064] Step S150 , performing pollution inversion analysis on the surface pollution signal to obtain inversion depth data, determining an operation cycle through the inversion depth data and a control amplification scheme, and performing pollution inversion operations based on the operation cycle to form a control enhancement field.

[0065] Specifically, the surface pollution signal is analyzed to obtain the inversion depth data. The reverse modeling technology is used to deduce the internal penetration of the pollutant from the pollution concentration and distribution information monitored on the surface. A physical model of pollution penetration is established, considering the diffusion, adsorption and chemical reaction of the pollutant, and the penetration equation uses the extended form of Fick's law where C is the pollutant concentration, D is the diffusion coefficient, and k is the reaction rate constant. The inversion process uses the finite element method to discretize the solution, and the street lamp surface is divided into triangular grid elements, and a local pollution distribution function is established in each element. The surface monitoring data and the theoretical model are matched by the least square fitting method, and the penetration depth distribution of the pollutant is solved. The inversion depth data includes the average penetration depth, the maximum penetration depth, the penetration unevenness and the penetration rate, etc. The multi-frequency excitation technology is used to enhance the resolution of the inversion analysis, and different frequency excitation signals correspond to different depth pollution information. The reliability analysis of the inversion results is established, and the confidence interval of the inversion depth data is calculated through the sensitivity analysis and the uncertainty propagation. The spatial interpolation processing is performed on the inversion depth data to generate the complete pollution depth distribution map of the street lamp surface.

[0066] The operation cycle is determined by the inversion depth data and the control amplification scheme. A multi-factor optimization method is used to consider the pollution penetration depth, the enhancement effect of the control amplification scheme and the cost-effectiveness of the cleaning operation. The operation cycle calculation formula T = a x D_avg + β x A_factor + γ x E_efficiency is established, where T is the operation cycle, D_avg is the average penetration depth, A_factor is the enhancement factor of the control amplification scheme, E_efficiency is the energy utilization efficiency, and a, β, γ are weight coefficients. According to the spatial distribution characteristics of the inversion depth data, the street lamp equipment is divided into different operation areas, and each area uses an independent operation cycle. The area with serious depth penetration uses a shorter operation cycle, and the area with lighter pollution uses a longer operation cycle. The enhancement effect of the control amplification scheme is reflected by the system response multiple, and the higher the response multiple, the better the control effect, and the operation cycle can be appropriately extended. A dynamic adjustment mechanism for the operation cycle is established, and the cycle parameters are adjusted in real time according to the changes of the pollution condition and the fluctuations of the control effect. An economic analysis method is used to balance the cleaning effect and the operation cost, and the optimal operation cycle configuration is found. A hierarchical management system of the operation cycle is designed, and the equipment is divided into different grades according to the importance and the pollution sensitivity, and different grades use different operation cycles.

[0067] In some embodiments, the pollution inversion operation based on the job cycle execution forms a control enhancement field, including: converting the job cycle into a cycle vector space; obtaining a balance core point in the cycle vector space; using the balance core point as a starting point to perform cycle diffusion to form a core enhancement area; and performing boundary optimization on the core enhancement area to form a control enhancement field.

[0068] The job cycle is converted into a cycle vector space. The construction of the cycle vector space considers multiple characteristic dimensions of the job cycle, including key parameters such as cycle length, start time, execution intensity, and coverage range. A base coordinate system of the vector space is established, linearly independent characteristic parameters are selected as base vectors to ensure the completeness and orthogonality of the vector space. The representation form of the job cycle vector is T_vec=[T_length, T_start, T_intensity, T_coverage], where each component corresponds to the time length, start time, execution intensity, and spatial coverage range of the cycle respectively. Standardization processing is used to convert the cycle parameters of different dimensions into dimensionless vector components, eliminating the influence of dimension difference on vector operation. A metric function of the vector space is established to define the distance and angle calculation method between vectors, supporting geometric operation and topological analysis in the vector space. The main change direction in the vector space is identified through principal component analysis method to determine the parameter combination that has the greatest impact on cycle characteristics. The subspace division of the vector space is established, and the vector space is decomposed into several subspaces according to the similarity of cycle characteristics, each subspace corresponds to a specific cycle mode. Vector clustering technology is used to group cycle vectors to identify cycle vector sets with similar characteristics.

[0069] The balance core point in the cycle vector space is obtained. The balance core point is defined as a special position in the vector space that makes the overall performance of the system reach the optimal balance state, and the vector distribution around this point has symmetry and stability characteristics. The barycenter calculation method is used to determine the geometric center of the vector set, and the barycenter coordinates are calculated by weighted average G=Σ(w i ×T i ) / Σw i , where w i is the weight coefficient of the i-th cycle vector, and T i is the corresponding vector coordinate. A balance evaluation index is established to quantify the balance degree by calculating the variance and skewness of the vector distribution. The smaller the variance, the more concentrated the distribution, and the smaller the skewness, the more symmetrical the distribution. An iterative optimization method is used to search for the optimal balance core point, and the optimization calculation is performed with the minimum imbalance degree of the vector distribution as the objective function. The stability analysis of the core point is established, and the sensitivity of the core point position to the change of the vector distribution is studied by perturbation analysis method. A multi-resolution search strategy is used to perform global search on a coarse grid to determine the approximate position of the core point, and then perform local search on a fine grid to obtain the accurate position.

[0070] Periodic diffusion is performed using the equilibrium core point as the starting point to construct a core enhancement region. With the identified equilibrium core point as the diffusion center, the surrounding core enhancement region is constructed through a diffusion process in vector space. Periodic diffusion employs a mathematical model similar to physical diffusion, where the diffusion intensity decays with increasing distance. The diffusion equation is I(r) = I0 × exp(-r² / 2σ²), where I(r) is the diffusion intensity at distance r, I0 is the initial intensity at the core point, and σ is the diffusion parameter. Anisotropic diffusion characteristics are established, using different diffusion coefficients in different vector directions to reflect the differential influence of the periodic parameter in each dimension. The boundary of the core enhancement region is determined using the isointensity surface method, defined as the point where the diffusion intensity decreases to a specific ratio of the initial intensity. A time evolution model of the diffusion process is developed to describe the dynamic expansion and contraction of the enhancement region over time. The shape of the core enhancement region is described by an ellipsoid, with the ratio of the major and minor axes of the ellipsoid reflecting the relative importance of different parameter dimensions. The volume and surface area of ​​the enhancement region are calculated using numerical integration methods to quantify the spatial extent and impact of the enhancement region.

[0071] The core enhancement area is optimized to form a control enhancement field. The boundary optimization adopts the variational method, with the optimization goal of minimizing boundary energy and maximizing internal consistency. Establish boundary smoothness constraints, and eliminate jagged edges and discontinuities of the boundary through spline fitting or NURBS surface fitting technology. Use morphological operations to regularize the boundary, and use expansion and erosion operations to fill boundary holes and eliminate boundary burrs. The intensity distribution of the control enhancement field is described by a continuous function to ensure the continuity and differentiability of the field distribution in space. Establish a gradient calculation method for the field , identifying the directions and regions where field intensity changes most dramatically, which typically correspond to key locations for control effectiveness. Multi-scale decomposition techniques are used to hierarchically describe the enhanced field, characterizing the field distribution at different spatial scales. Boundary conditions for the enhanced field are established, applying appropriate boundary conditions to ensure its physical plausibility. Finite element methods are used to numerically solve the enhanced field, yielding a high-precision numerical solution for the field distribution.

[0072] Step S160 , determining a response amplification window by controlling the enhancement field, and performing control gradient processing based on the response amplification window to obtain a cleaning control instruction.

[0073] Specifically, the response amplification window is determined by controlling the enhanced field. The response amplification window is determined by using the field strength threshold analysis method. When the field strength of the control enhanced field exceeds the set threshold F_threshold, the area is marked as a potential amplification window. Establish a field strength gradient calculation system , through the gradient amplitude The boundary regions with the most severe field strength changes are identified, which usually correspond to critical positions of the response amplification effect. The scattered high-intensity regions are aggregated into continuous window ranges using connectivity analysis techniques, and independent window regions are identified using the connected domain labeling method in morphological operations. The geometric characteristics of the response amplification window include window center position, coverage area, shape factor, and boundary curvature parameters. An evaluation index of window importance is established, considering factors such as the average field strength, field strength variance, and spatial continuity in the window. The identified multiple response amplification windows are prioritized, and the processing order of the windows is determined based on the field strength integral value and the influence range. A dynamic adjustment mechanism for the window is established to real-time correct the position and range of the window according to the time evolution of the control enhancement field. A multi-scale analysis method is used to identify response amplification windows at different spatial resolutions to ensure that amplification effects of different scales are captured.

[0074] Based on the response amplification window, the control gradient processing is performed to obtain the cleaning control instruction. The control gradient processing uses multi-directional gradient calculation to establish a two-dimensional or three-dimensional gradient field distribution within the window. The gradient calculation uses the finite difference method to perform numerical differentiation operations on the discretized window grid to calculate the gradient vector of each grid point. A gradient intensity classification system is established, and the regions within the window are divided into strong gradient zones, medium gradient zones, and weak gradient zones according to the gradient amplitude. The strong gradient zone corresponds to positions with severe pollution degree changes, which require high-intensity cleaning control strategies. The generation of the cleaning control instruction uses a hierarchical processing architecture, with the first layer being the cleaning mode selection, the second layer being the cleaning parameter setting, and the third layer being the cleaning timing arrangement. The optimal path of the cleaning action is determined based on the gradient direction information, and cleaning along the gradient direction can achieve the best pollution removal effect. A coding system for the control instruction is established to convert the gradient information into a machine executable digital instruction format. The cleaning control instruction includes key parameters such as cleaning position coordinates, cleaning intensity level, cleaning duration, and cleaning tool selection, and finally completes the intelligent control of the street lamp cleaning process.

[0075] In order to perform the street lamp cleaning process intelligent control method corresponding to the above-mentioned method embodiment, to realize the corresponding functions and technical effects. Referring to Figure 2 , Figure 2 The structure block diagram of a street lamp cleaning process intelligent control device 200 provided by an embodiment of the present application is shown. For ease of illustration, only the parts related to the present embodiment are shown. The street lamp cleaning process intelligent control device 200 provided by the embodiment of the present application comprises: The signal processing module 201 is configured to collect the surface pollution signal and the operation environment parameter of the street lamp device, convert the surface pollution signal to obtain pollution gain data through an adaptive sensing array, extract disturbance characteristic quantities from the operation environment parameter, and generate pollution driving response values by synergistically amplifying the pollution gain data and the disturbance characteristic quantities. The vibration processing module 202 is configured to construct a vibration induction chain based on the pollution driving response value, perform resonance detection on the vibration induction chain to obtain different frequency band vibrations, perform vibration conversion on the different frequency band vibrations to extract an energy collection point set, and construct a power enhancement matrix according to the energy collection point set. The dead zone optimization module 203 is configured to perform dead zone excavation analysis on the power enhancement matrix to identify an activatable dead zone range, determine a control gain position by using dead zone wake-up from the activatable dead zone range, and couple the control gain position with the pollution driving response value to determine an enhanced control node. The adaptive adjustment module 204 is configured to perform fault presetter adjustment based on the enhanced control node to determine an adaptive parameter, form a fault utilization chain by using the adaptive parameter through defect compensation control, perform synergistic fusion on the fault utilization chain to generate a control amplification scheme. The inversion control module 205 is configured to perform pollution inversion analysis on the surface pollution signal to obtain inversion depth data, determine an operation cycle by using the inversion depth data and the control amplification scheme, and perform pollution inversion operation based on the operation cycle to form a control enhancement field. The instruction generation module 206 is configured to determine a response amplification window by using the control enhancement field, perform control gradient processing based on the response amplification window to obtain a cleaning control instruction.

[0076] The street lamp cleaning process intelligent control device 200 described above can implement the street lamp cleaning process intelligent control method of the method embodiment described above. The optional items in the method embodiment described above are also applicable to this embodiment, and will not be described in detail herein. The remaining content of the embodiment of the present application can refer to the content of the method embodiment described above, and will not be described in detail herein.

[0077] As shown in Figure 3 The third embodiment of the present application further provides a computer device, which comprises a memory 301, a processor 302, and a computer program stored in the memory 301 and executable on the processor 302, and characterized in that the processor 302 implements the steps of the street lamp cleaning process intelligent control method according to the first embodiment of the present application when executing the program.

[0078] The purpose of the above embodiments is to exemplarily reproduce and deduce the technical solutions of the present application, and to completely describe the technical solutions, purposes and effects of the present application, so as to make the public more thoroughly and comprehensively understand the disclosed content of the present application, and not to limit the protection scope of the present application.

[0079] The above embodiments are also not exhaustive enumeration based on the present application, in addition to which, there can be a plurality of other embodiments not listed. Any substitution and improvement made without violating the concept of the present application is within the scope of protection of the present application.

Claims

1. A street lamp cleaning process intelligent control method, characterized in that: include: Collecting surface pollution signals and operating environment parameters of street lamp equipment, performing reverse encoding conversion on the surface pollution signals through an adaptive sensor array to obtain pollution gain data, extracting disturbance characteristic quantities from the operating environment parameters, and collaboratively amplifying the pollution gain data and the disturbance characteristic quantities to generate a pollution drive response value; constructing a vibration induction chain based on the pollution-driven response value, performing resonance detection on the vibration induction chain to obtain vibrations in different frequency bands, extracting an energy collection point set through vibration conversion of the vibrations in different frequency bands, and constructing a power enhancement matrix based on the energy collection point set; performing a dead zone mining analysis on the power enhancement matrix to identify an activatable dead zone range, determining a control gain position from the activatable dead zone range using dead zone wakeup, and coupling the control gain position with the pollution drive response value to determine an enhancement control node; performing fault presetter adjustment based on the enhanced control node to determine adaptive parameters, utilizing the adaptive parameters to form a fault utilization chain through defect compensation control, and performing collaborative integration on the fault utilization chain to generate a control amplification solution; Performing pollution inversion analysis on the surface pollution signal to obtain inversion depth data, determining an operation cycle based on the inversion depth data and the control amplification scheme, and performing pollution inversion operations based on the operation cycle to form a control enhancement field; A response amplification window is determined by the control enhancement field, and a control gradient process is performed based on the response amplification window to obtain a cleaning control instruction.

2. The method according to claim 1, characterized in that The collaboratively amplifying the pollution gain data and the disturbance characteristic quantity to generate a pollution drive response value includes: extracting a pollution gain coefficient from the pollution gain data; Constructing a gain-disturbance correlation curve using the disturbance characteristic quantity and the pollution gain coefficient; Obtaining a stable amplification section in the gain-disturbance correlation curve; The response value of the stable amplification section is determined as the pollution-driven response value.

3. The method according to claim 1, characterized in that The energy collection point set extracted by performing vibration conversion through vibrations in different frequency bands includes: Dividing the vibrations in different frequency bands into high-frequency blocks and low-frequency blocks; Projecting an energy conversion channel from the high-frequency block toward the low-frequency block; Tracking the coordinates of power enrichment points on the energy conversion channel to form an enrichment point group; The point with the strongest power in the enrichment point group is selected to form an energy collection point set.

4. The method according to claim 1, wherein The determining of a control gain position by using dead zone wake-up from the activatable dead zone range includes: The activatable dead zone range is subjected to response flow isolation identification to generate a flow isolation termination zone; Gain potential is calculated from the termination area of ​​the flow cutoff to form a gain multiplier; Obtaining the gain multiple and generating a continuous gain space through response fitting; The continuous gain space is used to perform boundary extraction to determine the control gain position.

5. The method according to claim 1, wherein The forming of a fault utilization chain by defect compensation control using the adaptive parameters includes: establishing a parameter time series from the adaptive parameters; Implementing defect marking on the parameter time series to form a compensation time node; Using the compensation time node as a dividing line, the parameter time series is divided into a pre-compensation period and a post-compensation period; Comparing the control characteristics of the pre-compensation stage and the post-compensation stage constitutes a fault utilization chain.

6. The method according to claim 1, characterized in that The performing of the pollution inversion operation based on the operation cycle to form a control enhancement field includes: Converting the operation cycle into a periodic vector space; Obtaining a balanced core point in the periodic vector space; Using the equilibrium core point as a starting point to perform periodic diffusion to form a core enhancement area; The core enhancement region is subjected to boundary optimization to form a control enhancement field.

7. The method according to claim 2, characterized in that The constructing of a gain-disturbance correlation curve using the disturbance characteristic quantity and the pollution gain coefficient includes: Identifying disturbance fluctuation characteristics according to the disturbance characteristic quantity to determine a construction range, wherein the disturbance fluctuation characteristics include disturbance amplitude, disturbance frequency and disturbance phase; Tracking the gain change process along the construction range based on the pollution gain coefficient to form a gain trajectory; Obtaining characteristic coordinates of each associated point in the gain trajectory; A gain-perturbation correlation curve is constructed from the correlation strength ranking of the feature coordinates.

8. The method according to claim 4, characterized in that The step of performing response flow isolation identification on the activatable dead zone range to generate a flow isolation termination zone includes: Performing reverse excitation injection on the boundary of the activatable dead zone range to generate a response echo; Performing a retrospective analysis on the attenuation trajectory of the response echo to determine the impedance transition point; Performing a response blocking process on the impedance jump point to form a flow blocking area; A flow blocking termination region is generated by defining the terminal boundary of the flow blocking region.

9. An intelligent control device for street lamp cleaning process, characterized in that: include: A signal processing module is configured to collect surface pollution signals and operating environment parameters of the street lamp equipment, perform inverse encoding conversion on the surface pollution signals through an adaptive sensor array to obtain pollution gain data, extract disturbance characteristic quantities from the operating environment parameters, and collaboratively amplify the pollution gain data and the disturbance characteristic quantities to generate a pollution drive response value; a vibration processing module, configured to construct a vibration induction chain based on the pollution-driven response value, perform resonance detection on the vibration induction chain to obtain vibrations in different frequency bands, perform vibration conversion on the vibrations in different frequency bands to extract an energy collection point set, and construct a power enhancement matrix based on the energy collection point set; a dead zone optimization module, configured to perform dead zone mining analysis on the power enhancement matrix to identify an activatable dead zone range, determine a control gain position from the activatable dead zone range using dead zone wakeup, and couple the control gain position with the pollution drive response value to determine an enhanced control node; an adaptive adjustment module, configured to adjust a fault presetter based on the enhanced control node to determine adaptive parameters, use the adaptive parameters to form a fault utilization chain through defect compensation control, and perform collaborative integration on the fault utilization chain to generate a control amplification solution; an inversion control module, configured to perform pollution inversion analysis on the surface pollution signal to obtain inversion depth data, determine an operation cycle based on the inversion depth data and the control amplification scheme, and perform pollution inversion operations based on the operation cycle to form a control enhancement field; The instruction generation module is used to determine the response amplification window through the control enhancement field, and perform control gradient processing based on the response amplification window to obtain the cleaning control instruction.

10. A computer device, characterized in that: The method comprises a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and implement the method according to any one of claims 1 to 8 when executing the computer program.

Images