857 results about "Linear relationship" patented technology

The invention discloses a ground motion intensity measure optimization method for seismic response prediction. The method includes the following steps: (1) building a building damage database; (2) optimizing ground motion intensity measures; (3) constructing a machine learning model; and (4) predicting damage states. According to the method, a combination of optimal ground motion intensity measures is selected by an elastic-net algorithm, wherein the efficiency and practicability in a current selection criterion are updated based on the goodness of fit and regression coefficient of the elastic-net model, thereby avoiding an assumption that intensity measures and structure response need to obey a logarithmic linear relationship and the limitation of intensity measure dimensions in traditional methods. By using the combination of optimal ground motion intensity measures as input variables, the prediction accuracy can be remarkably improved, and meanwhile the number of samples required is reduced.

The invention discloses a landslide disaster early warning system and method, and belongs to the technical field of geological disaster monitoring, and the method comprises the steps: carrying out the omnibearing monitoring of a target region, and collecting monitoring data; the method comprises the following steps: preprocessing collected monitoring data to obtain monitoring processing data; analyzing the monitoring processing data, and performing feature extraction to obtain a target feature vector; constructing a landslide disaster risk assessment model by adopting logistic regression and a decision tree; acquiring historical monitoring data according to a large amount of historical data and landslide cases, and training the model by using the historical monitoring data; real-time monitoring data are collected and input into the landslide disaster risk assessment model, the landslide risk level of the target area at the current moment is judged, and a corresponding early warning prompt is given; the method comprehensively considers the linear and non-linear relationship of the data, can accurately evaluate the landslide risk grade of the target area, and gives out graded early warning prompts. Compared with an existing scheme, the method has the advantage that the monitoring data source, the data processing precision and the model prediction capability are remarkably improved.

The invention, which belongs to the technical field of wireless communication, provides a position-guided channel knowledge graph construction method comprising the following steps: constructing an unmanned aerial vehicle auxiliary communication scene comprising a ground base station, and constructing a channel knowledge graph model; designing a back propagation neural network, training the back propagation neural network by using enough training samples corresponding to a ground base station j so as to learn a recessive nonlinear relationship between the position of the unmanned aerial vehicle and a channel gain, and representing a channel knowledge graph model as a dominant derivable function of the position of the unmanned aerial vehicle and the channel gain; and introducing a transfer learning algorithm, and performing fine tuning on the trained back propagation neural network of the ground base station j by using a limited sample corresponding to the target ground base station j 'to obtain a back propagation neural network of the target ground base station j', thereby obtaining a channel knowledge graph of the target ground base station j '. The method is a classical optimization method suitable for unmanned aerial vehicle trajectory optimization and deployment problems.

The invention provides a microscope automatic focusing method and device and a computer readable storage medium, and the method comprises the steps: S1, capturing a laser reflection light spot image on a sample through a camera, so as to determine the focusing position of an objective lens; s2, integrating the light spots through threshold processing and morphological operation, analyzing the gradient of a longitudinal gray distribution map to determine the edge and orientation, and finally identifying the maximum contour and framing the maximum contour with a minimum bounding rectangle to determine the width of the light spots; and S3, according to the linear relation between the light spot width and the defocusing distance and by analyzing the semi-circle orientation and width change of the light spot, controlling the focusing motor to reach the focusing position. The distance and the direction between the objective lens and the focal plane are determined by acquiring the reflected light spot image of the laser irradiated on the sample through the microscope objective lens, and the focusing error caused by the nonlinearity of the spot size when the point-semicircle spot approaches the focal plane is avoided according to the linear relation between the spot width and the defocusing distance. And the focusing process is high in efficiency and precision.

The invention relates to the technical field of valve internal leakage identification, aims to solve the problem of low leakage rate prediction precision in valve internal leakage acoustic emission detection, and discloses a nuclear power valve internal leakage detection method based on acoustic emission signal time-frequency characteristics. Extracting the characteristics of the time domain, the frequency domain and the time-frequency domain of the leakage signal, analyzing the linear relation between the characteristics and the leakage rate of the valve, and predicting the leakage rate and the leakage grade. According to the invention, remote online detection can be carried out on the internal leakage state of the valve, meanwhile, the internal leakage rate of the valve is predicted, the safety of maintenance personnel is guaranteed, the scheme feasibility is high, the measurement is accurate, the applicability is high, the improvement of automation and intelligence levels is facilitated, and good practicability and safety benefits are achieved.

The invention discloses an SAR imaging method and system based on sparse and low-rank decomposition, and a readable storage medium. The imaging method comprises the following steps: collecting an SAR image and two-dimensional echo data, and preprocessing the two-dimensional echo data to obtain a target reflectivity; and based on the linear relationship between the two-dimensional echo data and the target reflectivity, decomposing the SAR image by using a preset method to obtain a sparse component and a low-rank component. A decomposition result is iteratively updated through a preset optimization algorithm, the precision is improved until errors converge, and a sparse and low-rank decomposition algorithm expression is obtained; and solving a preset SAR imaging model by using the sparse and low-rank decomposition algorithm expression, outputting an SAR imaging result, and finally outputting the SAR imaging result. According to the invention, the complexity of SAR imaging can be effectively reduced, and the imaging efficiency is improved. Even if the data sampling rate is reduced, the accuracy of image reconstruction can be ensured.

The invention discloses an intelligent biological reaction monitoring system and method, and belongs to the field of intelligent monitoring, and the system comprises a reaction stage dividing unit which is used for dividing and determining multiple reaction stages; the variable relation mining unit is used for determining a dominant variable and an auxiliary variable, mining a variable linear relation and a collaborative insensitive loss function, and supervising and training the soft monitoring module; the continuous monitoring acquisition unit is used for configuring a monitoring equipment group, performing continuous monitoring and determining reaction monitoring data; the data analyzing and screening unit is used for preprocessing the data, performing linear analysis and abnormal reaction positioning and determining effective monitoring data; the game equilibrium decision-making unit is used for performing game equilibrium decision-making on the abnormal reaction identifier and determining a reaction adjustment strategy; and the reaction monitoring management unit is used for responding to the adjustment strategy, carrying out feedback monitoring analysis and reaction monitoring management, improving the biological reaction monitoring precision, realizing real-time online monitoring and carrying out intelligent regulation and control aiming at abnormal conditions.

The invention discloses a sensor fault diagnosis method, which comprises the following steps of: reading acquired data of a sensor in real time, and recording the data as a sensor reading at the current moment; judging whether the obtained sensor reading is within a preset threshold range or not; calculating sensor reading increments of adjacent time points, constructing increment distribution histograms of a plurality of time points, counting frequencies of different increment intervals, extracting increment distribution characteristic values, judging a ratio relation of the increment distribution characteristic values, and performing moving average processing and regression analysis on all sensor readings to obtain a plurality of sensor reading values; calculating a regression coefficient R to evaluate the linear relation strength between the data; comparing the Rvalue with a preset regression coefficient value; and judging whether the regression coefficient R is smaller than the preset value. Accurate diagnosis and positioning of sensor faults are realized, the false alarm rate is reduced, the reliability of a detection result is improved, and the method has universality, can be suitable for various types of environmental sensors, and meets different environmental monitoring requirements.

The invention discloses a motor electric control brake system clamping force self-adaptive correction method based on deep learning, and belongs to the technical field of chassis control, the method fully considers the wear condition of a brake pad in the use process, introduces a deep learning neural network optimization control strategy for the nonlinear change problem caused by the wear of the brake pad, and corrects the clamping force of the motor electric control brake system. The complex nonlinear relation between the clamping force estimated by the rotation angle of the motor and the actual clamping force is modeled as predictable probability distribution, and self-adaptive correction is performed, so that the clamping force control is optimized under the brake pad abrasion condition, the dynamic change caused by the brake pad abrasion is effectively dealt with, the real-time performance of system control is enhanced, and the system reliability is improved. The prediction precision and the response speed of the system are greatly improved, the stability, the reliability and the safety of the brake system are comprehensively enhanced, and meanwhile, the situation that a large number of sensors are adopted for real-time clamping force monitoring is avoided.

The invention discloses a diesel generator turn-to-turn short circuit fault protection method and system based on multi-parameter characteristics, and relates to the technical field of intelligent fault diagnosis of power equipment, and the method comprises the steps: carrying out the feature extraction of operation state monitoring data, carrying out the integration according to a time sequence, and generating a real-time feature sequence set; performing spatio-temporal feature coding and fusion on the real-time feature sequence set to generate a multi-dimensional spatio-temporal feature map; mapping the comprehensive matching score set into fault confidence according to a linear relation, obtaining a fault confidence sequence, judging a fault level in combination with a characteristic amplitude overrun condition, and generating a fault diagnosis result set; and executing hierarchical response based on the fault diagnosis result set, generating a hierarchical protection action execution record, and generating a short-circuit fault protection report in combination with the fault diagnosis result set. According to the method, quantitative similarity analysis of real-time features and standard modes is realized, the fault probability is accurately evaluated through a multi-parameter weighted fusion mechanism, and the reliability and interpretability of a diagnosis result are improved.

The invention relates to the technical field of machine learning optimization, in particular to a machine learning optimization method based on a QUBO model and quantum annealing, and the method comprises the following steps: firstly, discretizing continuous parameters of the machine learning model into binary variables, and constructing a QUBO objective function; then optimizing the QUBO model by using a quantum annealing algorithm; and finally, solving a globally optimal solution by adjusting annealing parameters, wherein the annealing parameters comprise the initial temperature, the cooling coefficient and the number of iterations. The QUBO model can discretize continuous variables (including parameters such as weights and offsets) in models such as AR, SVM and CNN in machine learning into binary variables, so that a nonlinear relation is better processed, and the calculation complexity in the training process is reduced. Through the QUBO model, the regularization item can be better controlled, and the problem of overfitting is avoided. Meanwhile, the QUBO model can be solved in parallel through quantum calculation or a simulated annealing algorithm, and the calculation efficiency under large-scale data is remarkably improved.

The invention discloses a navigation control method of an articulated vehicle and a related device, and relates to the field of automatic driving control, and the method comprises the following steps: obtaining navigation information of the articulated vehicle in real time, and calculating a steering curvature error based on the navigation information of the articulated vehicle and a preset discrete error tracking model. An expected value of the steering curvature is calculated based on the steering curvature error and a reference value of the steering curvature. And calculating the expected value of the hinge angle according to the expected value of the steering curvature. And generating a steering driving instruction to control a steering motor to drive the hinge angle to change into an expected value of the hinge angle. The discrete error tracking model is used for describing a linear relation between a pose tracking error vector in a discrete form and a steering curvature error, and the pose tracking error is composed of an X-axis tracking error, a Y-axis tracking error and a course angle tracking error. The motion state model of the articulated vehicle is linearized, and the control precision of curve tracking of the articulated vehicle is improved.

The invention discloses a fan unit blade root ultrasonic pretightening force monitoring method, and relates to the technical field of fan units, and the method comprises the following specific steps: carrying out ultrasonic fastening stress coefficient calibration on a fan unit blade root bolt material, applying axial tension to a plurality of sample bolts made of the same material, and determining the blade root ultrasonic pretightening force of the fan unit under each tension value; according to the method for monitoring the ultrasonic pretightening force of the blade root of the fan unit, the transmission speeds of ultrasonic waves in the sample bolt in different stress states are measured by utilizing ultrasonic detection equipment, and the accuracy of monitoring data can be remarkably improved; according to the method, the ultrasonic fastening stress coefficient of a bolt material is calibrated, high-precision real-time monitoring of the pretightening force of the bolt is realized by utilizing the linear relation between the ultrasonic wave spreading in the bolt and the stress, the influence of friction force on torque measurement is eliminated, and the misjudgment of pretightening force attenuation caused by temperature and vibration factors is also avoided.

The invention relates to the technical field of electric power fault detection, and discloses a fault monitoring system and method based on an electric power acquisition terminal, and the method comprises the steps: constructing a non-linear relation between the power consumption and the temperature through analyzing the historical power consumption data of a user and the weather temperature; the method comprises the following steps: dividing a user electricity consumption day into a plurality of time periods, and setting a time interval threshold value; and in each time period, data are acquired through the electric power and temperature acquisition devices, an electric quantity and temperature scatter diagram is constructed, and nonlinear data fitting is carried out. Thirdly, the similarity between the temperature nonlinear relational expressions of different time periods is calculated, the time periods are divided into sets, and a temperature retrieval set is established; and searching a corresponding electric quantity relation set by matching the temperature nonlinear relational expression of the user in the current time period, and judging whether electric power fluctuation and fault exist or not according to the priority ranking and the relation similarity.

The invention belongs to the technical field of detection, and discloses a nonlinear error correction method based on error change trend partition calibration for an eddy current displacement sensor. The method comprises the steps of firstly analyzing a corresponding relation between a displacement value and output voltage, and drawing an output characteristic curve of a sensor; distinguishing an error change trend according to a local extreme value, carrying out interval partitioning on an error change curve, training each partition, constructing a partition training set, constructing a correction model covering the full range of the sensor, and realizing error correction of high-precision measurement of the eddy current displacement sensor; and finally, the measured displacement value is obtained through the linear relation between the output voltage and the displacement value. According to the method, the linearity error, the sensitivity error, the return stroke error and the zero drift error after correction are greatly reduced, and the method is simple, easy to operate and high in real-time performance. Through reasonable segmentation, the complexity of curve fitting is reduced, the correction efficiency of the non-linear error of the eddy current sensor is improved, and the measurement precision is improved.

An optimization method and system based on low-rank regression of historical iteration data are provided. The method includes: obtaining historical iteration data including change data of physical parameters to be optimized and result deviation data obtained through simulation by using a forward model; and iteratively updating matrix coefficients and the physical parameters to be optimized continuously through low-rank linear regression based on the change data of the physical parameters to be optimized and the result deviation data, and evaluating updated physical parameters to be optimized, until an iteration stop condition is met. A low-rank matrix approximation that represents a small signal linear relationship between the change data of the physical parameters and the result deviation data in the historical iteration is construct and used to solve the optimization problem, which effectively reduces the computational complexity of the optimization process and improves the computational efficiency of the optimization process.

The invention provides an intelligent interpolation method for laser radar data missing, and the method comprises the steps: firstly obtaining long-time-sequence laser radar data, carrying out the preliminary screening and processing of the data through a quality control program, and then carrying out the classification of possible missing data. According to the method, linear function interpolation is executed on the condition of partial distance gate data missing caused by weather influence, values of missing data points are quickly and effectively estimated by utilizing a linear relation, and the continuity of the data can be well recovered in a scene that the data change is relatively stable. And for the condition of full-distance layer data missing in the time sequence caused by blade shielding and the like, a bidirectional recursion time sequence interpolation model based on LSTM is adopted, the advantages of a deep learning model in the aspect of processing time sequence data are fully played, data in a missing time period are accurately filled through historical data learning and bidirectional recursion prediction, and the accuracy of time sequence interpolation is improved. Therefore, the integrity and availability of the laser radar data are effectively improved.

The invention discloses a landslide risk and risk evaluation method suitable for a tectonic active area, and belongs to the field of geological disaster risk evaluation, and the risk evaluation method is characterized in that risk evaluation is superposed with vulnerability evaluation. The risk evaluation method comprises the following steps: calculating a deterministic coefficient value of each evaluation factor, and weighting to a grid layer of each evaluation factor to generate a deterministic coefficient grid layer of the evaluation factor; importing the deterministic coefficient layer into a logistic regression model, and generating a landslide risk evaluation result based on a logistic regression-deterministic coefficient coupling model; the vulnerability evaluation method is an analytic hierarchy process. By coupling a deterministic coefficient model and a logistic regression model, a linear relationship of multiple evaluation factors is effectively processed, and nonlinear contribution of a single evaluation factor to landslide development is captured, so that the landslide risk evaluation precision of a geological environment complex region and a structure active region is improved; and the evaluation result can more objectively and truly reflect the landslide risk of the active fault zone and the nearby area.

The invention relates to the technical field of power equipment fault diagnosis, in particular to a power transformer fault diagnosis method and system based on an improved graph convolutional network, and the method comprises the steps: obtaining a power transformer fault data set, and carrying out the preprocessing; an improved graph convolution network structure is constructed, and calculation of a graph convolution layer is improved through a self-circulation and adjacent matrix normalization method; dividing the power transformer fault data set into a training set and a test set, and performing sparse processing on an adjacent matrix; initializing the structure and parameters of the graph convolutional network; and training the graph convolutional network model by adopting a back propagation algorithm, and inputting an optimized result into a test set for verification to obtain a power transformer fault diagnosis result. According to the invention, the accuracy and efficiency of fault diagnosis of the power transformer are improved, the similarity information of the historical data and the current sample is fully utilized, and the defects that the complex nonlinear relationship is difficult to capture and the historical data is not fully utilized in the existing method are overcome.

The invention discloses a building carbon emission habit analysis method, device and equipment and a storage medium, and relates to the technical field of electric digital data processing. According to the method, the total energy consumption, such as the power consumption and the fuel combustion amount, of the building is obtained from resource purchasing which is easy to inquire, the total energy consumption, such as the power consumption and the fuel combustion amount, of the building can be obtained simply, and then each energy consumption device is separated to a thermal power station and a nuclear power station and to household appliances, cooking benches and the like of a residential building. The linear relation between the starting-up duration and the total carbon emission of each device is analyzed through linear regression, meanwhile, the size of a linear regression coefficient reflects the carbon emission of the device in unit time, and the larger the linear regression coefficient is, the larger the carbon emission of the device in unit time is. The carbon emission per unit time of each device can also be obtained from linear regression, and necessary data support is provided for the fields of subsequent large-scale carbon emission research and the like.

The invention discloses a carbon emission prediction method and device based on the relation between capture variables and a medium, and relates to the technical field of traffic carbon emission prediction. The method comprises the following steps: based on a preset traffic carbon emission influence factor set, calculating the correlation between the traffic carbon emission influence factor set and the traffic carbon emission, and carrying out preliminary screening to obtain a preliminary variable candidate pool; performing multi-collinearity detection on the preliminary variable candidate pool and updating the preliminary variable candidate pool to generate a screened variable candidate pool; performing nonlinear relation analysis on the screened variable candidate pool through an XGBoost algorithm, and screening to obtain a key variable sequence; inputting the key variable sequence and the traffic carbon emission into a Transform model, carrying out feature embedding and position coding, and generating a time sequence vector; and respectively predicting traffic carbon emission levels under different levels for the time sequence vectors. According to the method, the scientificity of the variable screening process and the reliability of model construction are ensured, and then the prediction capability of the model is enhanced.

The invention discloses an inductance material electromagnetic performance prediction method and system based on machine learning, and the method comprises the steps: collecting inductance performance index data, carrying out the data preprocessing of an inductance material performance index optimization data set, and carrying out the visualization of the data set, and obtaining the linear relation between the variables of the quantized inductance material performance index optimization data set. Through inductance material magnetic conductivity prediction, saturated magnetic field intensity prediction and inductance material electromagnetic performance prediction, inductance material electromagnetic performance characterization reference data are finally obtained. According to the method, the potential nonlinear relationship among the inductance raw materials is mined by utilizing the powerful calculation learning capability of the machine learning model, so that the raw material selection and formula design of inductance products with different specifications are accurately guided.

The invention discloses a machining product quality online detection precision evaluation method based on multi-factor fusion. The method comprises the following steps: firstly, determining machining process input factor characteristics of online detection precision evaluation and an output machining precision grade; secondly, historical samples of the factor characteristics are analyzed, an atomization sample plate set is generated, and a random fuzzy sample plate set is constructed; an online input feature sample is converted into a random fuzzy to-be-detected set, then the random fuzzy sample set is matched with the random fuzzy to-be-detected set, and reliability distribution describing the nonlinear mapping relation between the multi-environment-factor features and the detection precision is obtained. And finally, fusing the reliability distribution, determining the precision grade of online detection of the quality of the machined product according to a fusion result, and judging whether the product is qualified or not. According to the method, a complex non-linear relationship can be processed, reliability distribution related to the detection precision can be efficiently obtained, and the online detection precision can be evaluated more accurately in practical application.

The present application relates to the technical field of storage, and provides a method and apparatus for determining an optimal read voltage parameter of a flash memory, and a flash memory device. The method comprises: on the basis of a linear relationship model between a read retry voltage offset value and a bit flip count, acquiring a new read retry voltage offset value corresponding to a minimized predicted bit flip count of the flash memory; determining whether data can be successfully read under the new read retry voltage offset value; if yes, adding the new read retry voltage offset value to the header of a read retry offset table to form a new read retry offset table; and on the basis of the new read retry offset table, updating the linear relationship model by means of an adaptive learning model. By means of the efficient and adaptive voltage parameter determination method, the present application can improve the success rate of data recovery and reduce performance loss.

The invention discloses a climate mode pre-estimation downscaling correction method based on generative deep learning, and the method comprises the steps: recognizing the global climate mode hydro meteorological variable pre-estimation deviation through an unsupervised generative adversarial network, continuously reducing the adversarial loss through the adversarial training of a generator and a discriminator, and improving the accuracy of the adversarial training. Therefore, the trained generative deep learning network has the capability of processing a complex non-linear relationship so as to achieve the comprehensive optimum of downscaling and deviation correction. The method can be used for correcting the estimated average deviation of a multi-mode variable (rainfall, temperature and the like) set and improving the resolution ratio of the multi-mode variable set. The embodiment of the invention shows that the method can effectively estimate and downscale a multi-model set, and has a better downscaling effect than common quantile mapping and a convolutional neural network.

The invention provides an engineering operation vehicle load analysis method which comprises the following steps: predicting the change trend of the pressure of a hydraulic cylinder according to the descending speed and the load capacity of a fork, establishing a pressure prediction model through a support vector machine algorithm, and when the predicted pressure exceeds a safety threshold value, controlling the opening degree of a cushion valve in advance to realize predictive control of the pressure; a fuzzy rule base is established by combining the loading capacity of a forklift, the descending speed of a pallet fork and the pressure of a hydraulic cylinder, fuzzy self-adaptive control over the opening degree of a cushion valve is achieved, the opening degree of the cushion valve is automatically adjusted according to different working conditions, and it is ensured that the pressure of a hydraulic system is stabilized within a safety range. Establishing a nonlinear mapping relation among the loading capacity, the descending speed of the pallet fork and the opening degree of the buffer valve, training and optimizing a neural network model, achieving intelligent control over the opening degree of the buffer valve, and automatically adjusting the opening degree of the buffer valve according to the loading capacity and the descending speed of the pallet fork which are collected in real time.

The invention discloses a submarine sea mountain terrain inversion method based on a removal and recovery method, and the method comprises the steps: removing the influence of a sea mountain on a boundary based on a water depth forward modeling formula through the removal and recovery method by employing water depth forward modeling gravity abnormality and a vertical gravity gradient, and obtaining a high-precision sea mountain boundary terrain; and high-precision inversion is carried out on the sea and mountain main body by combining a GGM method. According to the method, the influence of sea and mountain topographic relief on boundary topography inversion in the linear relation is removed, the nonlinear relation in the linear relation is captured through deep learning, meanwhile, the water depth of a sea and mountain body is obtained through inversion of high-precision ship survey water depth points in combination with a GGM, and finally the high-precision submarine topography of the complete sea and mountain is obtained.

The invention discloses a vegetation index time sequence prediction method and device based on deep learning, and the method comprises the following steps: obtaining and processing historical period vegetation indexes and environmental factor data of a preset time period, and obtaining NDVI, EVI and kNDVI data as vegetation indexes, environment factor data including temperature, rainfall, evapotranspiration, wind speed, relative humidity and soil humidity; by determining the intensity and direction of the linear relation between the environmental factors and the vegetation indexes, the environmental factors with high correlation with the vegetation indexes are selected; a CNN-BiLSTM-AM model comprising an input layer, a CNN layer, a BiLSTM layer, an AM layer and an output layer is constructed to predict a vegetation index time sequence, environmental factors with high correlation serve as input data, and NDVI, EVI and kNDVI vegetation indexes serve as output prediction targets. The technical process is simple and clear, and the accuracy of vegetation index time sequence prediction is greatly improved.

The invention provides a precision compensation method and device of a magnetic encoder based on linear displacement detection, and relates to the field of precision compensation of magnetic encoders, and the method comprises the steps: carrying out the equal-interval division of a total reference range of the movement of a moving part, which is determined by a reference sensor, obtaining a reference output value comprising a preset number of range calibration points and each range calibration point; obtaining a difference value corresponding to each range calibration point of the moving part; a preset algorithm is adopted to calculate the target reference output value, the first reference output value, the second reference output value, the first difference value and the second difference value of any detection point where the moving part is located, and a target error is obtained; determining a target movement output value of the magnetic encoder at the detection point based on the target error, the target reference output value and a linear relation corresponding to two adjacent range calibration points before and after the detection point; according to the method, precision compensation is carried out on the magnetic encoder, so that the moving output value in the subsequent independent application process of the magnetic encoder is more accurate.

The invention discloses an industrial park load analysis method based on an optimized K-means clustering algorithm, and the method comprises the steps: collecting the power load data of an industrial park, including the internal data of the total power consumption of equipment, the time-sharing power consumption, the power consumption peak value, the energy consumption of the equipment and the like, and the external data of weather data, electricity price market information, policies and regulations and the like; the method comprises the following steps of: performing unified standardization processing on data, including cleaning and normalizing the data, and filling by adopting a KNN method aiming at the problems of missing values and errors in the data, so as to avoid the problem of low baseline load analysis accuracy caused by errors in data collection; afterwards, a KPCA nonlinear dimension reduction technology is adopted to process a nonlinear relation between the power demand of the production line and external data, the complexity of a model is reduced by simplifying a data set, more nonlinear features in the external data are extracted, and the rule of the power demand of the industrial park can be better mastered; the problem that a power demand side management method is short in processing a non-linear relation is solved.