CN-121998249-A - New energy station operation simulation assessment method and system based on comprehensive early warning

CN121998249ACN 121998249 ACN121998249 ACN 121998249ACN-121998249-A

Abstract

The invention provides a new energy station operation simulation and assessment method and system based on comprehensive early warning, and relates to the field of new energy power generation, wherein the method comprises the steps of collecting operation and fault data, constructing a parameter propagation network to identify fluctuation abnormality and generating a risk propagation path; the method comprises the steps of determining turning points according to fluctuation change rates, dividing adjustment and disturbance time periods, constructing a staged evolution scene, injecting control instructions and interference signals, evaluating adjustment and disturbance rejection capability based on response and deviation of a destabilization track and a safety boundary, and generating an assessment result. The method realizes early warning and targeted evaluation of the running risk of the new energy station.

Inventors

WU DALI
ZHENG JIANHU
WANG JUNFENG
MU HAO

Assignees

陕西华电新能源发电有限公司

Dates

Publication Date: 20260508
Application Date: 20260126

Claims (10)

1. The new energy station operation simulation assessment method based on comprehensive early warning is characterized by comprising the following steps of: Collecting real-time operation data and historical fault data of a new energy station; Extracting fluctuation characteristics of each operation parameter before failure occurrence from historical failure data, calculating correlation coefficients among the fluctuation characteristics, and constructing a parameter propagation network according to the correlation coefficients; Inputting real-time operation data into a parameter propagation network, identifying initial fluctuation abnormality, extracting a propagation sequence related to the initial fluctuation abnormality in the parameter propagation network, and generating a risk propagation path; Extracting fluctuation change rate of operation parameters on a risk propagation path from historical fault data, determining fluctuation turning points, and dividing an adjustment period and a disturbance period based on the fluctuation turning points; Constructing a staged evolution scene according to the risk propagation path, injecting a correction control instruction in an adjustment period, injecting a quantized interference signal in a disturbance period, and recording a response track of the correction control instruction and a destabilization track of the quantized interference signal; calculating the dynamic deviation between the response track and the preset safety boundary to obtain an adjustment capability score, calculating the dynamic deviation between the instability track and the preset safety boundary to obtain an anti-interference capability score, and combining the adjustment capability score and the anti-interference capability score to generate an assessment result.
2. The method of claim 1, wherein extracting fluctuation features of each of the operating parameters before occurrence of the fault from the historical fault data, calculating correlation coefficients between each of the fluctuation features, and constructing the parameter propagation network based on the correlation coefficients comprises: Positioning a fault trigger time from the historical fault data, intercepting a historical data segment with preset duration forwards by taking the fault trigger time as a termination time, extracting fluctuation amplitude and fluctuation period of each operation parameter in the historical data segment, and combining to form fluctuation characteristics of each operation parameter; Calculating a time delay correlation coefficient between fluctuation characteristics of any two operation parameters, and screening the operation parameter pairs with the time delay correlation coefficient reaching preset strength; Judging an operation parameter which firstly fluctuates and an operation parameter which later fluctuates in the operation parameter pair, setting the operation parameter which firstly fluctuates as a propagation source node, setting the operation parameter which later fluctuates as a propagation target node, and taking a time delay correlation coefficient between the propagation source node and the propagation target node as an inter-node correlation coefficient; and constructing a parameter propagation network according to the propagation source node, the propagation target node and the inter-node correlation coefficient, wherein the nodes of the parameter propagation network are operation parameters, the edges are the connection of the propagation source node to the propagation target node, and the weight of the edges is the inter-node correlation coefficient.
3. The method of claim 1, wherein inputting the real-time operational data into a parameter propagation network, identifying an initial fluctuation anomaly, extracting a propagation sequence associated with the initial fluctuation anomaly in the parameter propagation network, and generating a risk propagation path comprises: inputting the real-time operation data into a parameter propagation network, extracting real-time fluctuation characteristics of each operation parameter in the real-time operation data, performing deviation calculation on the real-time fluctuation characteristics and fluctuation characteristics of corresponding nodes in the parameter propagation network, marking the nodes with deviation exceeding a preset deviation threshold as initial fluctuation anomalies, and recording the fluctuation directions of the initial fluctuation anomalies; Searching downstream nodes along a directed edge according to the node positions in the initial fluctuation abnormal parameter propagation network, extracting real-time fluctuation characteristics of operation parameters corresponding to the downstream nodes, calculating fluctuation directions corresponding to the real-time fluctuation characteristics of the downstream nodes, and screening the downstream nodes with the fluctuation directions consistent with the fluctuation directions of the initial fluctuation abnormal parameters to form a candidate propagation node set; Extracting the correlation coefficient between nodes in the parameter propagation network of each node in the candidate propagation node set, calculating the path correlation strength from the initial fluctuation anomaly to each node in the candidate propagation node set, screening nodes with the path correlation strength greater than a preset strength threshold and corresponding propagation paths, and forming a propagation sequence related to the initial fluctuation anomaly; A risk propagation path is generated from propagation levels and path correlation strengths of nodes in the propagation sequence.
4. The method of claim 1, wherein extracting a fluctuation rate of change of the operational parameter on the risk propagation path from the historical fault data, determining a fluctuation turning point, and dividing the adjustment period and the disturbance period based on the fluctuation turning point comprises: Constructing sliding observation windows with different time lengths, extracting fluctuation data of operation parameters on a risk propagation path, calculating the mean value and variance of the fluctuation data, selecting an optimal observation window according to a minimum criterion of the variance, and resampling the operation parameters on the risk propagation path by using the optimal observation window to obtain a fluctuation sampling sequence; carrying out layered decomposition on the fluctuation sampling sequence to obtain a plurality of layers of fluctuation components, calculating energy distribution values of the plurality of layers of fluctuation components, selecting a target fluctuation component according to the energy distribution values, and recombining the target fluctuation component to generate a fluctuation change sequence; Calculating the variation between adjacent sampling points of a fluctuation variation sequence to obtain the fluctuation variation rate of the operation parameters on the risk propagation path, segmenting the fluctuation variation rate, identifying the variation direction of the fluctuation variation rate, and marking the sampling point with the reverse variation direction as a fluctuation turning candidate point; Calculating the fluctuation amplitude ratio and the duration of the fluctuation turning candidate points to form fluctuation morphological characteristics, and screening to obtain the fluctuation turning points according to the fluctuation morphological characteristics; according to the fluctuation turning point, the fluctuation change sequence is segmented, the time period of the fluctuation change rate from negative to positive is divided into adjustment time periods, and the time period of the fluctuation change rate from positive to negative is divided into disturbance time periods.
5. The method of claim 1, wherein constructing a staged evolution scenario from a risk propagation path, injecting a correction control command during an adjustment period, injecting a quantized interference signal during a disturbance period, and recording a response trajectory of the correction control command and a destabilization trajectory of the quantized interference signal comprises: collecting state variables of adjacent control nodes on a risk propagation path, calculating the mean value and fluctuation amplitude of the state variables, determining an evolution boundary, and dividing the evolution boundary into an adjustment evolution scene and a disturbance evolution scene based on a fluctuation turning point; Generating a correction control instruction in the regulation evolution scene, calculating the fluctuation frequency and the fluctuation amplitude of the correction control instruction, and injecting the correction control instruction with the minimum fluctuation frequency and the fluctuation amplitude meeting the preset fluctuation range into adjacent control nodes according to time intervals to obtain correction input signals of the control nodes; Collecting state response data corresponding to a correction input signal of a control node, calculating the change rate and the change acceleration of the state response data, constructing a response track, and calculating the acting time of a correction control instruction based on the response track; Dividing a fluctuation period based on action time length in a disturbance evolution scene to generate a quantized interference signal, calculating amplitude distribution of the quantized interference signal, and injecting the quantized interference signal with the amplitude distribution meeting a preset distribution range into adjacent control nodes to obtain interference input of the control nodes; Based on the interference input of the control node, collecting state fluctuation data, calculating the amplitude change rate and the frequency offset of the state fluctuation data, constructing a destabilization track, and outputting a response track and the destabilization track.
6. The method of claim 5, wherein generating a quantized interference signal within the perturbation evolution scene based on dividing the period of fluctuation by the duration of action comprises: Segmenting the action duration according to a preset proportion coefficient to obtain a plurality of fluctuation periods, and calculating the average value period of each fluctuation period; Generating a cosine waveform reference signal in the mean value period, taking the peak position and the trough position of the reference signal as characteristic points, and constructing a time sequence of the characteristic points; Nonlinear transformation is carried out on the time sequence of the characteristic points to generate a plurality of groups of distorted characteristic point sequences, and segmentation reconstruction is carried out on the reference signals based on the distorted characteristic point sequences to obtain a plurality of groups of distorted interference signals; Calculating the frequency spectrum distribution and the energy distribution of the distorted interference signals, and constructing a fluctuation feature vector of the distorted interference signals; And carrying out matching degree calculation on the fluctuation feature vector and a preset disturbance template, and selecting a distortion interference signal with the highest matching degree as a quantized interference signal.
7. The method of claim 1, wherein calculating the dynamic deviation of the response trajectory from the preset safety boundary to obtain the adjustment capability score, calculating the dynamic deviation of the destabilization trajectory from the preset safety boundary to obtain the immunity capability score, and combining the adjustment capability score with the immunity capability score to generate the assessment result comprises: extracting an upper limit threshold value, a lower limit threshold value and a standard frequency value of a preset safety boundary to generate safety boundary parameters; Segmenting the response track according to the sampling time sequence, calculating the fluctuation deviation of each segment of response track and the safety boundary parameter, and constructing a fluctuation amplitude deviation sequence and a fluctuation frequency deviation sequence; Calculating the accumulated time length exceeding the safety boundary parameter in the fluctuation amplitude deviation sequence to obtain amplitude out-of-limit time length, and calculating the accumulated amplitude deviating from the standard frequency value in the fluctuation frequency deviation sequence to obtain frequency offset; Converting the amplitude out-of-limit time length and the frequency offset according to a preset mapping coefficient to generate an adjustment capacity score; Segmenting the destabilization track according to a fluctuation period, calculating the fluctuation change rate of each segment of destabilization track and the safety boundary parameter, and constructing a fluctuation rate sequence and a fluctuation acceleration sequence; Calculating the number of mutation points in the fluctuation speed sequence and the transition amplitude value in the fluctuation acceleration sequence, and converting the number of mutation points and the transition amplitude value according to a preset mapping coefficient to generate an anti-interference capacity score; and generating an assessment result according to the adjustment capability score and the disturbance rejection capability score by a preset combination weight.
8. A new energy station operation simulation assessment system based on comprehensive early warning, for implementing the method of any one of the preceding claims 1-7, characterized by comprising: the first unit is used for collecting real-time operation data and historical fault data of the new energy station; The second unit is used for extracting fluctuation characteristics of each operation parameter before the occurrence of the fault from the historical fault data, calculating correlation coefficients among the fluctuation characteristics, and constructing a parameter propagation network according to the correlation coefficients; the third unit is used for inputting the real-time operation data into the parameter propagation network, identifying the initial fluctuation abnormality, extracting a propagation sequence related to the initial fluctuation abnormality in the parameter propagation network and generating a risk propagation path; A fourth unit for extracting a fluctuation change rate of the operation parameter on the risk propagation path from the history fault data, determining a fluctuation turning point, and dividing the adjustment period and the disturbance period based on the fluctuation turning point; A fifth unit, configured to construct a staged evolution scenario according to the risk propagation path, inject a correction control instruction in the adjustment period, inject a quantized interference signal in the disturbance period, and record a response track of the correction control instruction and a destabilization track of the quantized interference signal; And a sixth unit, configured to calculate a dynamic deviation between the response track and a preset safety boundary to obtain an adjustment capability score, calculate a dynamic deviation between the destabilization track and the preset safety boundary to obtain an immunity capability score, and combine the adjustment capability score and the immunity capability score to generate an assessment result.
9. An electronic device, comprising: A processor; A memory for storing processor-executable instructions; Wherein the processor is configured to invoke the instructions stored in the memory to perform the method of any of claims 1 to 7.
10. A computer readable storage medium having stored thereon computer program instructions, which when executed by a processor, implement the method of any of claims 1 to 7.

Description

New energy station operation simulation assessment method and system based on comprehensive early warning Technical Field The invention relates to a new energy power generation technology, in particular to a new energy station operation simulation and assessment method and system based on comprehensive early warning. Background As the power generation duty ratio of new energy is continuously increased, safe and stable operation of new energy stations becomes increasingly important to the reliability of the power system. At present, a new energy station faces challenges such as changeable weather conditions, frequent equipment faults, complex control systems and the like, and an effective early warning mechanism and an assessment method are needed to ensure the stable operation of the new energy station. The traditional new energy station operation assessment is mainly based on static index evaluation, so that the dynamic response characteristic of the station under the complex operation environment is difficult to effectively simulate, and the adjustment and disturbance rejection capability of the station under the abnormal working condition cannot be comprehensively reflected. Most of existing new energy station assessment methods adopt single index evaluation, usually only concern about static deviation of parameters such as output power, voltage and the like, lack analysis of dynamic fluctuation characteristics of operation parameters and propagation rules thereof, and cannot predict possible fault risks and evolution paths, so that early warning is not timely and measures are not accurate. The traditional assessment method can not distinguish the challenge difference of the new energy station in different operation stages, particularly can not distinguish the period of time required to be regulated by the system and the period of time suffered from external disturbance, so that the assessment result can not truly reflect the coping capability of the station in different fault types and development stages, and the assessment result is lack of pertinence. The existing assessment system lacks a comprehensive assessment mechanism for dynamic response capability of a new energy station, cannot simultaneously consider the adjustment capability and the disturbance rejection capability of the station, cannot verify the comprehensive response capability of the station by simulating a real fault scene, so that a large gap exists between an assessment result and an actual running condition, and cannot provide effective guidance for station optimization. Disclosure of Invention The embodiment of the invention provides a new energy station operation simulation and assessment method and system based on comprehensive early warning, which can solve the problems in the prior art. In a first aspect of the embodiment of the present invention, a new energy station operation simulation and assessment method based on comprehensive early warning is provided, including: Collecting real-time operation data and historical fault data of a new energy station; Extracting fluctuation characteristics of each operation parameter before failure occurrence from historical failure data, calculating correlation coefficients among the fluctuation characteristics, and constructing a parameter propagation network according to the correlation coefficients; Inputting real-time operation data into a parameter propagation network, identifying initial fluctuation abnormality, extracting a propagation sequence related to the initial fluctuation abnormality in the parameter propagation network, and generating a risk propagation path; Extracting fluctuation change rate of operation parameters on a risk propagation path from historical fault data, determining fluctuation turning points, and dividing an adjustment period and a disturbance period based on the fluctuation turning points; Constructing a staged evolution scene according to the risk propagation path, injecting a correction control instruction in an adjustment period, injecting a quantized interference signal in a disturbance period, and recording a response track of the correction control instruction and a destabilization track of the quantized interference signal; calculating the dynamic deviation between the response track and the preset safety boundary to obtain an adjustment capability score, calculating the dynamic deviation between the instability track and the preset safety boundary to obtain an anti-interference capability score, and combining the adjustment capability score and the anti-interference capability score to generate an assessment result. Extracting fluctuation features of each operation parameter before failure occurrence from the historical failure data, calculating correlation coefficients among the fluctuation features, and constructing a parameter propagation network according to the correlation coefficients comprises the following steps: Positioning a fault trigge