194 results about "Local outlier factor" patented technology

The invention provides a water turbine anomaly detection and fault diagnosis method based on multi-sensor data fusion, and the method comprises the steps: collecting stress data, audio signals, vibration signals and working condition data during the operation of a water turbine, and carrying out the preprocessing of the collected data, including data cleaning, alignment and normalization; extracting time-domain, frequency-domain and time-frequency-domain features of the preprocessed stress data, audio signals and vibration signals, forming a high-dimensional feature vector in combination with the preprocessed working condition data, and performing dimensionality reduction on the high-dimensional feature vector by using principal component analysis; performing anomaly detection on the dimensionality-reduced high-dimensional feature vector according to a clustering algorithm, a local outlier factor algorithm and an anomaly detection algorithm constructed by an auto-encoder neural network, and generating a comprehensive anomaly score in combination with a preset weight to obtain an anomaly detection result; and when the water turbine is in an abnormal state, the fault type is diagnosed through the similarity between the dimensionality-reduced high-dimensional feature vector and the historical fault feature template.

A fault detection method for refrigeration units based on an improved deep learning model is provided, including the following steps: S1: obtaining operating parameters of a refrigeration unit in a normal operating state and in states with different fault types as data sets; S2: detecting local outliers in the data set by using a local outlier factor algorithm and removing the local outliers, and then expanding the data set by using adaptive synthetic sampling; S3: normalizing the data set; S4: constructing a fault detection model; and S5: inputting the parameters of the tested refrigeration unit into the fault detection model, and judging whether the tested refrigeration unit has a fault and the type of the fault.

The invention relates to the field of switching power supply high-side current detection, in particular to a switching power supply high-side current detection method, which comprises the following steps of: acquiring current signals to form a sequence, acquiring an intrinsic mode by using empirical mode decomposition, analyzing each mode to construct a difference amplitude-frequency index, and calculating a noise index in combination with a similarity index and heterogeneity. According to the method, the de-noising window is intelligently adjusted, and the local outlier factor is utilized to detect the anomaly, so that accurate current anomaly identification is finally realized, the detection precision and the system intelligence level are remarkably improved, the resource utilization is optimized, and a powerful guarantee is provided for stable operation of a power system.

A method for monitoring temperature-induced deformation of components in an intelligent moxibustion robot based on infrared spectroscopy includes the following steps. Infrared spectral images and frequency spectra of a target component of the intelligent moxibustion robot at different detection points are acquired. Motion influence confidence factors for each detection position are constructed based on frequency differences between peaks and troughs in the frequency spectrum. The box-counting method is used to obtain scale-relationship graphs of infrared spectra for all detection positions. The overall light absorption difference index of the target component is determined according to the scale-relationship graphs. By combining the overall light absorption difference index with motion influence confidence factors, local outlier factors (LOF) for each detection position in the thermal data sequence are calculated using a LOF anomaly detection algorithm. Finally, a temperature deformation risk of the target component is evaluated based on a thermal alarm threshold.

The invention relates to an input data quality evaluation method for photoetching process window analysis. The method comprises the following steps: S1, obtaining standardized data; s2, adopting an RANSAC regression algorithm to obtain a robust regression residual analysis confidence coefficient; s3, obtaining the confidence of the isolation tree based on the abnormal score of the isolation tree; s4, adopting a local outlier factor method to obtain local outlier factor confidence; s5, integrating the isolation tree confidence coefficient and the local outlier factor confidence coefficient to obtain an unsupervised confidence coefficient; s6, obtaining a Gaussian confidence coefficient; and S7, integrating the robust regression residual analysis confidence coefficient, the unsupervised confidence coefficient and the Gaussian confidence coefficient to obtain a comprehensive confidence coefficient, and obtaining an input data quality evaluation result for photoetching process window analysis. Compared with the prior art, the method has the advantages of improving the input data quality of photoetching process window analysis and the like.

The invention relates to a wind turbine generator set anomaly detection method considering multi-unit state similarity. Firstly, an evaluation index system is constructed, and design parameters and historical operation data are analyzed. Then data acquisition and preprocessing are carried out, an SCADA system is used for acquiring data, and denoising is carried out in combination with a local outlier factor algorithm and wavelet transform; and then qualitative evaluation is carried out, a cloud model is constructed to describe data uncertainty, a membership degree is calculated, and a cluster is dynamically adjusted. And then comprehensive evaluation is carried out, index weights are determined through an analytic hierarchy process, and scores are calculated in combination with membership degrees. And finally, performing anomaly detection, setting a threshold value, combining a dynamic clustering method, and taking a quantile of a sliding time window and a similar unit prediction residual error as a real-time early warning threshold value. According to the method, the accuracy and reliability of wind turbine generator anomaly detection are improved.

The invention discloses a battery pack abnormal monomer detection method and system based on a local outlier factor algorithm. The method comprises the following steps: sampling and collecting voltage data and current data of each battery monomer in a battery pack by adopting a sliding window method to obtain a monomer voltage matrix; respectively obtaining a first voltage characteristic F1 and a second voltage characteristic F2 according to the monomer voltage matrix U; calculating a voltage change consistency feature F3 of each cell in the sliding window according to the cell voltage matrix U and the current data of the cell; integrating the first voltage feature F1, the second voltage feature F2 and the voltage change consistency feature F3 into feature points, and integrating and mapping the feature points into data points; and performing anomaly detection on the data point set by adopting a local outlier factor (LOF) algorithm to obtain an abnormal data point which is an abnormal single battery in the battery pack. According to the invention, the accuracy of the abnormal cells in the battery pack can be improved, and the abnormal condition type can be judged.

The invention discloses a DC bus interference anomaly detection method based on a dynamic local outlier factor algorithm. The DC bus interference anomaly detection method is especially suitable for real-time operation state monitoring of a complex power system. Aiming at the limitation that a traditional anomaly detection method is high in calculation complexity and limited in detection precision in high-dimensional time series data processing, optimization is carried out on the basis of an outlier detection theory, a dynamic mode updating algorithm is provided, the algorithm evaluates outlier characteristics of time series data through local density, and the detection precision of the time series data is improved. The global calculation overhead is obviously reduced, and the real-time performance and the detection efficiency of the system are improved. Compared with a traditional method, the method consumes manual intervention, is high in detection precision and strong in early warning, can be widely applied to the field of health assessment and intelligent operation maintenance of power equipment, and shows a good engineering application prospect.

The invention relates to a malicious website identification method and system based on IP address feature analysis, and belongs to the technical field of network security, and the method comprises the steps: collecting network flow data of an IP address through a network flow log, a firewall log and a DNS query record, and carrying out the preprocessing; constructing an IP dynamic graph structure by taking the IP addresses as nodes and the communication relationship between the IP addresses as edges according to the preprocessed network flow data; the IP dynamic graph structure is coded through a graph attention network, the complex relation between nodes is captured, the time sequence characteristics of IP address behaviors are captured in combination with a long and short-term memory network, and a behavior mode of the nodes is constructed; and calculating a reconstruction error and a local outlier factor LOF of the node according to a behavior pattern of the node, calculating an abnormal score of the IP address, and when the abnormal score is higher than a dynamic abnormal detection threshold, performing deep analysis by combining domain name characteristics, webpage content characteristics and network access behavior characteristics of the website, and identifying the malicious website.

The invention provides an industrial anomaly detection and root localization method and system based on data driving, and the method comprises the steps: carrying out the cleaning, feature extraction and normalization processing of original data collected in an industrial production process, and constructing a feature space; based on a local anomaly factor LOF and a mahalanobis distance MD method, jointly detecting local anomaly and global anomaly, and identifying an abnormal working condition; extracting space and time correlation characteristics of the abnormal variables through Pearson correlation weighting and Granger causal test to form a space-time correlation matrix; constructing an abnormal causal network based on the matrix, and tracing an abnormal root and a propagation path through depth-first search and abnormal propagation intensity evaluation; and finally, dynamic optimization of the anomaly detection and diagnosis method is realized based on parameter self-adaption and model incremental learning. According to the method, the anomaly detection accuracy and the anomaly traceability interpretation capability can be effectively improved, and the intelligent level and the self-adaptive capability of data processing are enhanced.

The invention discloses a power distribution network grounding line selection analysis and fault location method based on signal feature identification. The method comprises the following steps: extracting a fault signal of a monitoring point; adopting an AGCN-TCN model to carry out fault identification and classification on the fault signals; local outlier factors of the monitoring points are calculated, fault signals of all the monitoring points are decomposed according to sliding windows, sliding window frequency domain assignments obtained after decomposition serve as frequency domain fault indexes, fault position judgment comprehensive indexes are constructed based on the local outlier factors and the frequency domain fault indexes, and the positions of fault points are judged through the fault position judgment comprehensive indexes; according to the method, the AGCN-TCN model is innovatively constructed for the distributed power distribution network, feature extraction and analysis of fault signals are realized from the angles of space and time, and decision errors caused by faults of single monitoring equipment are avoided; according to the method, a fault position judgment comprehensive index is defined, the fault position is judged from the double angles of the time domain and the frequency domain, and inaccuracy of a single method is avoided.

The invention discloses an automatic conductive ink screen printing quality detection method based on a big data model. The method comprises the following steps: acquiring hyperspectral image data, three-dimensional shape data and real-time impedance distribution data of a conductive ink screen printing layer; performing local outlier factor optimization processing on the hyperspectral image data, and extracting a pixel-level spectral feature vector; performing vertical weight function Gaussian regression filtering on the three-dimensional shape data to generate a topological smooth curved surface coordinate set; optimizing the real-time impedance distribution data based on the conductivity distribution matrix, and outputting a noise reduction conductivity matrix; constructing an association graph and a node feature tensor based on the detection key points, and generating silk-screen multi-modal data; through modal interaction, features are extracted in stages, and cross-modal joint features are extracted through multi-stage joint clustering optimization; and a task space separation head module is input, and a quality evaluation result is output through classification and regression branches. According to the method, multi-modal data deep fusion, noise robustness improvement, cross-modal feature accurate extraction and efficient real-time detection are realized.

The invention relates to the technical field of electrical digital data processing, in particular to an airborne navigation data cleaning method based on multi-source sensing data, which comprises the following steps: acquiring parameter data of airborne navigation at a plurality of historical time nodes, constructing a parameter sequence, and constructing a multi-scale window by taking any parameter data in the parameter sequence as a center; and calculating an abnormal score of any parameter data under any scale window by using an improved local outlier factor algorithm, judging the parameter data of which the mean value of the abnormal scores of any parameter data under all scale windows is greater than a set threshold value as abnormal data, and cleaning the abnormal data. The problem of misjudgment or missed judgment in the data cleaning process is solved.

The invention provides an electronic current transformer operation data calculation system, and relates to the technical field of data processing, and the system comprises a data processing module which is used for carrying out the sliding time window scanning of an equipment dependence adjacency matrix, calculating the local reachable density under a dynamic neighbor topological distance through a local outlier factor algorithm, if it is detected that the harmonic total distortion rate gradient enters a statistical hypothesis test rejection domain or the winding temperature rise rate exceeds a material thermal aging critical threshold, extracting a multi-dimensional abnormal feature vector set; the risk quantitative index generation module is used for generating a risk quantitative index group from the multi-dimensional abnormal feature vector set through a pre-trained fault propagation probability model; and the processing instruction set generation module is used for establishing a multi-objective optimization function based on the risk quantification index group, performing non-dominated sorting evolution on a processing instruction chromosome by adopting an NSGA-II genetic algorithm, screening a dynamic optimization instruction sequence matched with a real-time load fluctuation coefficient through a Pareto frontier solution set, and generating a structured processing instruction set. The operation reliability of the power system is improved.

The invention discloses a data optimization storage method based on artificial intelligence, and relates to the field of data storage, and the method comprises the steps: collecting the storage rate data of a storage unit and the transmission rate data of a data transmission unit; carrying out anomaly detection on the storage rate data by adopting a density-based local anomaly factor LOF algorithm; adopting a multi-model combination prediction method based on SARIMA and LSTM to obtain a predicted transmission rate sequence at the next moment; calculating a residual error between the acquired transmission rate data and the predicted transmission rate sequence; constructing a deep reinforcement learning model based on near-end strategy optimization PPO; and constructing a multi-target reward function according to the storage rate abnormal point sequence and the transmission rate residual sequence, training a deep reinforcement learning model by maximizing the target reward function, and generating an optimal data transmission rate control strategy. Aiming at unbalanced utilization of storage resources in the prior art, the application balances the utilization of the storage resources.

The invention discloses a multi-human-body-target behavior recognition method based on millimeter-wave radar. The method comprises the following steps: firstly, acquiring human body behavior data by deploying a millimeter wave radar, and obtaining a multi-dimensional parameter signal of human body behaviors by adopting 4D-Fast Fourier Transform (4D-FFT); secondly, a high-dimensional matrix signal is converted into a point cloud matrix through a cell average constant false alarm rate (CAFAR) detection algorithm (CA-CFAR) and geometric space coordinate conversion, so that the complexity of a radar signal is effectively reduced; secondly, further compressing the point cloud matrix through a density-based spatial clustering (DBSCAN) algorithm and a local outlier factor (LOF) algorithm based on space and speed information to remove dynamic interference and static interference in the environment, and then realizing segmentation of the multi-target point cloud matrix through a Hungary algorithm to obtain a pure single-target point cloud matrix; and finally, realizing accurate recognition of human body behaviors through a neural network based on the multi-dimensional radar point cloud. According to the invention, through radar point cloud and deep learning, the complexity of radar signals is reduced and accurate recognition of human body behaviors is realized at the same time.

The invention discloses an abnormal traffic flow detection method and system based on a dynamic graph, and is applied to the field of abnormal traffic flow detection. The method comprises the following steps: dividing a city into a plurality of sub-regions through city road network data, and obtaining vehicle trajectory data to construct a traffic flow tensor; constructing an approximate graph based on the traffic flow tensor and performing data enhancement to generate a local view and a global view of the traffic flow tensor; designing a comparative learning model with a time graph encoder to perform joint training on the multi-view data, and capturing space-time dependency features; and acquiring a real-time traffic flow embedding vector by adopting a sliding window, identifying an abnormal traffic mode through a local abnormal factor algorithm, and finally realizing the positioning of an urban abnormal traffic area. According to the method, through fusion of the time graph encoder and comparative learning, the accuracy of traffic anomaly detection in a complex road network environment is effectively improved, and reliable technical support is provided for urban traffic intelligent management.

The invention belongs to the technical field of photovoltaic power generation, and relates to a photovoltaic power generation power prediction method and device and a computer readable storage medium. Collecting meteorological parameters and photovoltaic power generation powers at a plurality of sampling moments to obtain photovoltaic power generation data sequences at the plurality of sampling moments, and calculating a correlation coefficient between each meteorological parameter and the photovoltaic power generation power; taking the K photovoltaic power generation data sequences as clustering centers, and clustering all the photovoltaic power generation data sequences to obtain K initial photovoltaic power generation data clusters; calculating a local outlier factor and a weight of each photovoltaic power generation data sequence in each initial photovoltaic power generation data cluster; constructing a photovoltaic power generation power prediction probability distribution model based on the weights of all the photovoltaic power generation data sequences and Bayesian linear regression model posterior distribution; the parameters of the photovoltaic power generation power prediction probability distribution model are optimized by using the photovoltaic power generation data sequence, the target photovoltaic power generation power prediction model is obtained, and the accuracy of the photovoltaic power generation power prediction result is improved.

The invention discloses a battery consistency detection method, which analyzes voltage time sequence data of battery cells in a battery cluster by introducing a local outlier factor algorithm, and can accurately identify outlier battery cells with abnormal voltage behaviors in charging and discharging processes. Compared with a traditional index threshold value judgment method, due to the fact that the local outlier factor algorithm and the reinforcement learning mechanism are combined, the optimal algorithm parameters are automatically searched, the method does not depend on manually-set experience threshold values any more, and the detection stability and reliability are improved. Meanwhile, compared with a statistical feature method, the method can more effectively capture local abnormal features in the battery cell voltage data, is not limited by overall data distribution hypothesis, has higher tolerance to abnormal data, and has smaller dependence on a judgment threshold. The method can effectively improve the precision of consistency detection of the battery cells in the battery cluster, provides a more reliable basis for subsequent battery cell charging, and further remarkably improves the overall capacity utilization rate of the battery cluster, prolongs the service life, and reduces the operation and maintenance cost.

The invention discloses a millimeter wave water surface flow velocity measurement radar CFAR detection method and system, and the method comprises the steps: eliminating an outlier in advance through a local outlier factor, carrying out the estimation of the quartile distance of a sample through the screened data, obtaining a cut-off threshold value of the sample, and finally obtaining a background clutter power estimation value through the maximum likelihood estimation. According to the method, the selection of the quartile optimization truncation threshold is introduced, and the reference window sample is screened in advance by using the local outlier factor, so that the robustness of the quartile and the truncation threshold is enhanced; according to the method, the environmental adaptability of the algorithm is enhanced, the target detection performance of the algorithm in a multi-target and clutter edge environment is effectively improved, and a feasible new scheme is provided for target detection of the millimeter wave water surface flow velocity measurement radar.

The invention discloses a power distribution switch terminal communication state analysis method and system, and belongs to the technical field of power distribution network communication. The method comprises the following steps: acquiring dual-source parameters of a terminal and a communication link, and preprocessing to form a time sequence data sequence; an improved local outlier factor algorithm is combined with spatio-temporal clustering to identify abnormal communication events; extracting an analysis sequence containing event context based on the dynamic window; respectively performing communication state evaluation on a terminal side and a link side through a multi-task deep learning model, and outputting state classification and severity score; and when the evaluation results on the two sides conflict, intelligent decision making is performed by adopting a multi-level arbitration mechanism based on an evidence theory and integrating factors such as confidence difference, a historical trend and topological association, so that accurate responsibility determination of a fault source is realized. The method solves the problems that the traditional threshold alarm is high in false alarm rate and the fault source cannot be positioned, and remarkably improves the accuracy and operation and maintenance efficiency of communication state diagnosis.

The invention discloses a water quality monitoring intelligent early warning method and system based on multi-dimensional data analysis, and belongs to the technical field of water quality monitoring. The method comprises the following steps: processing missing values and noise through a dynamic weighted data filling algorithm to obtain complete water quality data; carrying out multi-parameter collaborative dimensionality reduction by adopting an improved t-SNE algorithm, and mapping high-dimensional water quality parameters to a three-dimensional feature space; constructing a dynamic watershed partition model based on a Delaunay triangulation network and DEM data; calculating a spatial anomaly score through a weighted local outlier factor; performing time anomaly detection in combination with STL (Standard Template Library) decomposition and a Transform model; multi-scale fluctuation detection is realized by using wavelet packet decomposition; fusing space, time and parameter dimensions to construct a three-dimensional judgment matrix, and generating a comprehensive anomaly score; and triggering a multi-level early warning mechanism according to a scoring result. The system can monitor the water quality change in real time, and gives out early warning in time when abnormity occurs, so that the water quality safety is guaranteed.

The invention discloses a battery life prediction method based on density outlier factor elimination. The method comprises the following steps: firstly, carrying out charge-discharge circulation on a lithium battery, analyzing a voltage-current change curve of the lithium battery in a charge-discharge stage, and selecting a plurality of HIs capable of representing battery degradation from the voltage-current change curve; then, the correlation degree between each HI sequence and the battery capacity is quantitatively analyzed, and meanwhile, standardization processing is carried out on the extracted HI sequences; then, on the basis of a local outlier factor algorithm, outliers are removed in sequence, redundancy and information bias are reduced, and the data density is improved; and finally, the cleaned HI sequence is used as an input and output sequence of the model, so that the effectiveness and accuracy of the battery life prediction model are improved. According to the method, the case library is constructed by using the actual operation battery attenuation data, and powerful guidance is provided for an operator to find the battery life prediction accuracy.

The invention discloses a big data analysis-based regional business environment evaluation and analysis system and method. The method comprises the steps of collecting multi-source customer service data; identifying abnormal growth points in the data by calculating the approximation degree between adjacent subdivision time nodes; calculating a local outlier factor (LOF) for sampling data, and evaluating the abnormal degree of a data point; meanwhile, in the scheme, based on the LOF and the abnormal growth evaluation value, the comprehensive abnormal evaluation score of each datum is calculated, the abnormal degrees of different data points are compared more visually, and the analysis operability is enhanced; and comparing the comprehensive anomaly evaluation score with a preset threshold value, identifying a result and generating a report to provide decision support, thereby facilitating optimization of the overall operation efficiency in the region.

The invention discloses a bacterial liquid capsule production control method and system based on the Internet of Things, and belongs to the technical field of production control. Abnormal values are recognized and eliminated in a first data set by combining a local outlier factor algorithm and a K-means outlier threshold value, the abnormal values in the data are effectively recognized, the real effectiveness of the data is improved, and the production efficiency of the bacterial liquid capsule is improved. And fuzzy division is carried out on a historical normal data set through a clustering algorithm to obtain a clustering center value of each clustering cluster, each data in the second data set is clustered, and a corresponding abnormal value is repaired through a mean value in the corresponding cluster, so that data missing after the abnormal value is removed is avoided, the continuity of the data is ensured, and the production efficiency is improved. And the effectiveness of production control is further improved.

The invention discloses a method for repairing reliability of a whole fracturing pump based on dynamic fault analysis, which comprises the following steps of: S1, acquiring operation data of the whole fracturing pump, and preprocessing and storing the operation data by utilizing edge computing equipment; s2, extracting short-term features and long-term features of the operation data by adopting an adaptive segmentation method of a dynamic window, and carrying out anomaly detection by combining an isolated forest and a local anomaly factor method; s3, constructing a fracturing whole pump fault propagation model, updating the fault propagation probability by using a Markov chain Monte Carlo method, and generating a fault propagation influence matrix; s4, evaluating a fault mode by adopting a dynamic fault tree-Markov chain conjoint analysis method to form a fault influence degree matrix; s5, potential failure point prediction is carried out, and a fault evolution trend chart is generated; and S6, constructing a three-layer optimal scheduling framework by adopting an optimal scheduling method of a game theory, and generating an optimal repair strategy. According to the method, dynamic fault analysis, game theory optimization and the like are combined, and accurate prediction and intelligent repair of the whole fracturing pump are achieved.

The invention is suitable for the technical field of virtual power plants, and provides a virtual power plant data processing method and system. According to the method, data acquisition is carried out on a plurality of application ends, and data preprocessing is carried out through a plurality of edge data nodes, so that a plurality of pieces of effective acquisition data are obtained; transmitting the plurality of effective acquisition data to a cloud end, and performing data integration and standardization through the cloud end to generate standard integrated data; neglecting defect attribute data, and adopting a C4.5 decision tree algorithm to perform classification mining on defect-free attribute data to obtain classification mining data; and performing outlier detection of data points on the classified mining data, calculating a plurality of local outlier factors, and performing abnormity checking and processing. According to the method, cloud edge collaboration can be carried out, data preprocessing is carried out through edge data nodes, the computing pressure of a cloud end is effectively reduced, the data processing efficiency is improved, a C4.5 decision tree algorithm is adopted, data classification mining is carried out, outlier detection and anomaly checking and processing are carried out, and accurate use of data is ensured.

The invention relates to the technical field of radiotherapy data analysis, provides a tumor radiotherapy abnormity early warning system based on a clustering algorithm, and aims to improve the abnormity detection capability and early warning precision of patient health data in a radiotherapy process. According to the system, by introducing a local outlier factor and an adaptive bandwidth optimization technology, density weighted calculation of a Mean Shift clustering algorithm is optimized, an internal mode in tumor patient data can be accurately identified, and the accuracy of preliminary clustering is improved; in the enhanced clustering process, an entropy regularization optimization P3D cutting method is adopted, strategy network updating of a PPO enhanced learning algorithm is further optimized, clustering features are dynamically adjusted, and the adaptability and accuracy of the system in a dynamic environment are enhanced; according to the system, reinforcement learning and clustering analysis technologies are combined, high-precision and high-real-time support is provided for abnormal early warning in the radiotherapy process, and the intelligent level and reliability of patient monitoring are remarkably improved.

The invention relates to the technical field of monitoring and alarming, in particular to an online monitoring method and system for a coking coal detection chamber, and the method comprises the steps: collecting parameter data of a current time node in the coking coal detection chamber and parameter data in a plurality of historical set time windows; calculating an abnormal score of the parameter data of the current time node under any historical set time window by using an improved local outlier factor algorithm to obtain abnormal scores of the parameter data of the current time node under all historical set time windows, and when the mean value of the abnormal scores in all the historical set time windows is greater than a set abnormal threshold value, determining that the current time node is abnormal, and triggering an alarm device to perform early warning. The problem that an existing algorithm is not high in detection accuracy is solved.

The invention discloses a river channel water body dissolved oxygen prediction method based on information construction, and relates to the technical field of environmental protection information processing. An LSTM neural network algorithm based on information construction is proposed to predict the dissolved oxygen of the river water body, a density-based local outlier factor algorithm and a linear interpolation method are used to carry out abnormal value detection and data filling, SG filtering is used to denoise the data, and the data are calculated according to the day and night volatility of the dissolved oxygen and the daily change rule of the dissolved oxygen. According to the method, the difference between the maximum value and the minimum value of the daily dissolved oxygen and the daily time are quantified into numerical values to serve as the characteristics of a time sequence, the LSTM neural network model is finally constructed, the problem that due to the fact that the day and night fluctuation of the dissolved oxygen is large, the prediction precision is limited can be effectively solved, and the prediction precision is effectively improved.