Hydropower centralized control center computer monitoring method, system and device

Abstract

The application discloses a hydropower centralized control center computer monitoring method, system and device, relates to the technical field of power system automation monitoring, and is specially aimed at a small hydropower station group operation scene, provides a hydropower centralized control center computer monitoring method, and improves the method on the basis of the monitoring system technology of the existing hydropower centralized control center, carries out long-time interval steady-state analysis and short-time interval transient-state analysis through setting a first data processing channel and a second data processing channel, and determines the equipment operation state based on a joint criterion, can effectively identify the equipment abnormality of multiple time scales, can simultaneously monitor the steady-state deviation abnormality and the transient-state impact abnormality in the equipment operation, avoids the abnormality missed report caused by a single time scale, and thus improves the accuracy and timeliness of hydropower station equipment monitoring.

Description

Technical Field

[0001] This application relates to the field of power system automation monitoring technology, and in particular to computer monitoring methods, systems and devices for hydropower centralized control centers. Background Technology

[0002] The hydropower centralized control center is the core of the operation of a small hydropower station cluster. Its monitoring system undertakes functions such as remote monitoring, data acquisition, and command issuance. It realizes centralized resource scheduling by building an interaction channel between upper and lower computers, which is crucial to power grid security and energy efficiency. Early systems were mostly based on the x86 architecture, relying on specific hardware and closed operating systems to form a stable but highly coupled ecosystem. Initially, it could meet basic data processing and control needs, but with technological iteration and increased security requirements, architectural defects and rigid data processing logic became apparent.

[0003] Hydropower station equipment anomalies exhibit characteristics across multiple time scales: bearing overheating caused by component wear and guide vane jamming require lengthy analysis, while vibration spikes caused by mechanical loosening require transient detection. However, existing systems use fixed time windows for processing; long time windows smooth out transient signals, leading to missed detections, while short time windows miss steady-state shifts, creating early warning blind spots.

[0004] In monitoring hydroelectric turbine units, long-term averages can mask the impact signals of mechanical loosening, while short-term analysis struggles to detect gradual temperature changes. This insufficient processing capability leads to the neglect of fault symptoms, hindering refined condition assessment and seriously threatening operational safety. A multi-timescale collaborative data processing mechanism is urgently needed to overcome these limitations. Summary of the Invention

[0005] In order to at least overcome the above-mentioned deficiencies in the prior art, the purpose of this application is to provide a computer monitoring method, system and device for hydropower centralized control center to solve the above problems.

[0006] Firstly, this application provides a computer monitoring method for a hydropower centralized control center, including... Acquire raw telemetry data from one or more devices deployed on the side of the hydropower station, configure timestamps, configure the corresponding monitoring type according to the preset device identification classification module, and obtain the raw data sequence; In response to obtaining the original data sequence, the first data processing channel and the second data processing channel are simultaneously started; The first data processing channel is configured to: preset a first time interval, obtain a first statistical feature quantity based on the original data sequence of any device within the first time interval; compare the first statistical feature quantity with a preset first threshold interval, and obtain a first-level warning signal if the threshold is exceeded. The second data processing channel is configured to: preset a second time interval, obtain subsequences within each window through a sliding window method to obtain a second statistical feature; compare the second statistical feature with a preset second threshold, and obtain a second-level warning signal if the threshold is exceeded; Based on the first-level warning signal and the second-level warning signal, and using a preset joint criterion module, the equipment operating status category is determined; Based on the determined equipment operating status category, and using a preset control strategy mapping table, control commands are obtained and sent to the equipment on the hydropower station side.

[0007] In one possible implementation, a third data processing channel is also synchronously started in response to obtaining the original data sequence; The third data processing channel is configured as follows: According to the monitoring type, a corresponding preset third time interval is obtained, and the length of any third time interval is within the length of the first time interval and the second time interval; Based on the original data sequence of any device within the third time interval, calculate the coefficient of variation to obtain the third statistical characteristic. When the third statistical feature is greater than the preset third threshold, it is determined that the device operating status is abnormal and the device operating status category is not normal.

[0008] In one possible implementation, the first data processing channel is configured to: preset a first time interval; obtain a first statistical feature quantity based on the original data sequence of any device within the first time interval; compare the first statistical feature quantity with a preset first threshold interval; and obtain a first-level warning signal if the threshold is exceeded, including: Based on the monitoring type of the original data sequence, a first time interval is set; The raw data sequence within the first time interval is obtained from any device at a set sampling period, and the validity is filtered to obtain the valid data sequence; Based on the effective data sequence, the arithmetic mean of the effective data sequence is calculated to obtain a first statistical characteristic quantity to characterize the steady state of the device's operating state over a long period of time; Based on the first statistical feature, a real-time comparison is performed within a preset first threshold range; If the first statistical feature is not within the preset first threshold range, it is determined that a steady-state shift anomaly has occurred. In response to a steady-state shift, a first-level warning signal is triggered.

[0009] In one possible implementation, the second data processing channel is configured to: preset a second time interval, which is shorter than the preset first time interval; obtain a subsequence within each window using a sliding window approach to obtain a second statistical feature; compare the second statistical feature with a preset second threshold; if the threshold is exceeded, obtain a second-level warning signal, including: Based on the first time interval, a second time interval shorter than the length of the first time interval is defined; Based on the original data sequence, the step size is set according to the length of the sampling period, and the second time interval is used as a sliding window to move and obtain a continuous data subsequence within each window; Based on the continuous data subsequences, the standard deviation of each continuous data subsequence is calculated to obtain a second statistical characteristic quantity, which characterizes the degree of data fluctuation of the device within a very short time window; Based on the second statistical feature, a real-time comparison is performed using a preset second threshold. If the second statistical feature is not within the preset second threshold range, it is determined that there is a transient impact anomaly. In response to the presence of a transient impact anomaly, a level 2 warning signal is triggered.

[0010] In one possible implementation, determining the equipment operating status category based on the first-level warning signal and the second-level warning signal, using a preset joint criterion module, includes: The preset joint criterion module is configured as follows: Upon receiving a Level 1 warning signal, it is determined that the first statistical feature is not within the first threshold range; Upon receiving a second-level warning signal, it is determined that the second statistical feature is greater than the second threshold. By taking whether the first-level warning signal and the second-level warning signal are received as conditions, a Boolean combination is performed to obtain four equipment operating status categories, including normal state, steady-state deviation anomaly, transient impact anomaly, or composite anomaly. The first-level warning signal and the second-level warning signal are input into the preset joint criterion module, which outputs the equipment operating status category.

[0011] In one possible implementation, the condition of whether a first-level warning signal and a second-level warning signal are received is used as a Boolean combination to obtain four equipment operating state categories, including normal state, steady-state deviation anomaly, transient impact anomaly, or combined anomaly, including: If neither the first-level warning signal nor the second-level warning signal is received, the state is considered normal. If a Level 1 warning signal is received but a Level 2 warning signal is not received, the steady-state offset is determined to be abnormal. If no Level 1 warning signal is received, but a Level 2 warning signal is received, it is determined to be a transient impact anomaly. If a Level 1 warning signal is received and a Level 2 warning signal is received, it is determined to be a composite anomaly. The first-level warning signal and the second-level warning signal are input into a preset joint criterion module, and the device operation status category is output. When outputting the status category, a confidence score is also output. Based on the first statistical feature, the second statistical feature, the first threshold interval, and the second threshold, the degree to which the first statistical feature deviates from the first threshold interval and the extent to which the second statistical feature exceeds the second threshold are calculated to obtain a confidence score, which characterizes the accuracy of the determined state category.

[0012] In one possible implementation, the method further includes updating the first threshold interval and the second threshold, including: From the data of the past set time period when the equipment operating status category was marked as normal, all first statistical features and second statistical features are obtained to obtain a sample set; Based on the sample set, calculate the mean and standard deviation of the first statistical feature within the set time period to obtain the updated first threshold interval; Based on the sample set, calculate the second statistical feature quantity within the set time period to obtain the updated second threshold; In response to the updated first threshold range and the updated second threshold being confirmed a second time by the operations and maintenance personnel, the current first threshold range and second threshold are obtained.

[0013] One possible implementation also includes: Based on the original data sequence, a time series model trained on historical data when the equipment operating status category was marked as normal was used to obtain the real-time residual variance. If the real-time residual variance exceeds the residual variance of the historical data when the device operating status category is marked as normal, and exceeds a set multiple for a continuous set sampling period, then the device operating status category is determined to be potentially abnormal, and a third-level warning signal is output.

[0014] Secondly, this application provides a computer monitoring system for a hydropower centralized control center, including a data acquisition unit, a data processing unit, a first data processing channel unit, a second data processing channel unit, a unit for determining equipment operating status, and an instruction issuing unit that are sequentially electrically connected. The data acquisition unit is configured to: acquire raw telemetry data from one or more devices deployed on the side of the hydropower station, configure timestamps, configure corresponding monitoring types according to the preset device identification classification module, and obtain raw data sequences; The data processing unit is configured to: in response to obtaining the original data sequence, simultaneously start the first data processing channel and the second data processing channel; The first data processing channel unit is configured to: preset a first time interval, obtain a first statistical feature quantity based on the original data sequence of any device within the first time interval; compare the first statistical feature quantity with a preset first threshold interval, and obtain a first-level warning signal if the threshold is exceeded; The second data processing channel unit is configured to: preset a second time interval, obtain subsequences within each window through a sliding window to obtain a second statistical feature; compare the second statistical feature with a preset second threshold, and if it exceeds the threshold, obtain a second-level warning signal; The device operation status determination unit is configured to: determine the device operation status category based on the first-level warning signal and the second-level warning signal, and based on a preset joint criterion module; The instruction issuing unit is configured to: obtain control instructions based on the determined equipment operating status category and a preset control strategy mapping table, and issue them to the equipment on the hydropower station side.

[0015] Thirdly, this application provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of any one of the methods for computer monitoring of a hydropower central control center.

[0016] In summary, the beneficial effects that this application can achieve are: This application provides a computer monitoring method, system, and device for a hydropower centralized control center. By setting up a first data processing channel and a second data processing channel to perform long-term steady-state analysis and short-term transient analysis respectively, and determining the equipment operating status based on a joint criterion, it can effectively identify equipment anomalies at multiple time scales. It can simultaneously monitor steady-state deviation anomalies and transient impact anomalies during equipment operation, avoiding missed anomalies caused by a single time scale, thereby improving the accuracy and timeliness of hydropower station equipment monitoring. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings: Figure 1This is a schematic diagram of the method steps in an embodiment of this application; Figure 2 This is a schematic diagram of the method flow of an embodiment of this application; Figure 3 This is a schematic diagram of the system structure according to an embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the accompanying drawings in this application are for illustrative and descriptive purposes only and are not intended to limit the scope of protection of this application. Furthermore, it should be understood that the schematic drawings are not drawn to scale. The flowcharts used in this application illustrate operations implemented according to some embodiments of this application. It should be understood that the operations in the flowcharts may not be implemented in sequence, and steps without logical contextual relationships may be reversed or implemented simultaneously. In addition, those skilled in the art, guided by the content of this application, may add one or more other operations to the flowcharts, or remove one or more operations from the flowcharts.

[0019] Furthermore, the described embodiments are merely some, not all, of the embodiments of this application. The components of the embodiments of this application described and illustrated herein can typically be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0020] Example 1 Please refer to the following: Figure 1 The above is a schematic diagram of the steps of the hydropower centralized control center computer monitoring method provided in the embodiment of the present invention. Further, the hydropower centralized control center computer monitoring method may specifically include the contents described in steps S1-S6.

[0021] Step S1: Obtain raw telemetry data from one or more devices deployed on the hydropower station side, configure timestamps, configure the corresponding monitoring type according to the preset device identification classification module, and obtain the raw data sequence.

[0022] Step S2: In response to obtaining the original data sequence, the first data processing channel and the second data processing channel are started simultaneously.

[0023] Step S3: The first data processing channel is configured to: preset a first time interval, obtain a first statistical feature quantity based on the original data sequence of any device within the first time interval; compare the first statistical feature quantity with a preset first threshold interval, and obtain a first-level warning signal if the threshold is exceeded.

[0024] Step S4: The second data processing channel is configured to: preset a second time interval, obtain subsequences within each window through a sliding window to obtain a second statistical feature; compare the second statistical feature with a preset second threshold, and if it exceeds the threshold, obtain a second-level warning signal.

[0025] Step S5: Based on the first-level warning signal and the second-level warning signal, and using the preset joint judgment module, determine the equipment operating status category, including one of four categories: normal state, steady-state deviation anomaly, transient impact anomaly, or composite anomaly.

[0026] Step S6: Based on the determined equipment operating status category and the preset control strategy mapping table, obtain control commands and send them to the equipment on the hydropower station side.

[0027] In practical applications, this embodiment of the application utilizes a dual-channel parallel processing structure to achieve multi-timescale collaborative analysis of the operating status of hydropower station equipment.

[0028] During the data acquisition phase, raw telemetry data from one or more devices deployed on the hydropower station side are acquired and timestamps are configured. The corresponding monitoring type is then configured according to the preset device identification classification module to obtain the raw data sequence. Acquiring raw telemetry data involves directly extracting real-time operating parameters from the sensors or data acquisition devices of the hydropower station side equipment. The hydropower control center computer receives the data uploaded by the devices via wireless transmission. After receiving the uploaded data, a timestamp is configured, which can be achieved through the system clock synchronization mechanism of the control center computer, ensuring that all data have a unified time reference. The corresponding monitoring type is then configured according to the preset device identification classification module. This can be achieved using rule matching based on device model or function type, such as querying the device identifier against a predefined device type database to determine the type of the corresponding monitoring parameter. This ensures data time synchronization and device specificity. The design of configuring monitoring types according to device identification classification allows for different processing based on the physical characteristics of different devices (as shown in Example 2, different time intervals are set), avoiding the weakening of specific device anomalies by a general processing mode.

[0029] When the raw data sequence is ready, the first and second data processing channels are started synchronously. This synchronization mechanism avoids the time delay introduced by sequential processing, thus enabling real-time response to the dynamic changes in the operating status of the hydropower station's equipment.

[0030] In the first data processing channel, a first time interval needs to be preset. This interval can be set to a fixed length or dynamically adjusted mode according to the operating characteristics of the equipment. For example, an hourly time window can be used for bearing temperature rise monitoring, while a day-level time window can be selected for component wear monitoring. Then, based on the original data sequence of any equipment within this interval, a first statistical characteristic is calculated. This can include, but is not limited to, calculating the arithmetic mean, weighted average, or moving average as the first statistical characteristic, thereby accurately reflecting the operating trend of the equipment over a long time scale. Afterward, this statistical characteristic is compared with a preset first threshold interval. The comparison can be implemented through a simple numerical comparison algorithm, or a dynamic threshold adjustment mechanism based on historical data can be introduced to adapt to changes in the operating environment of different equipment. If the statistical characteristic exceeds the threshold interval, a first-level warning signal is triggered. The core advantage of this design is that the long time interval can effectively smooth out short-term noise interference, highlight the slow changing trend of the equipment during long-term operation, and thus accurately identify gradual anomalies such as bearing temperature rise and component wear.

[0031] In the second data processing channel, a second time interval shorter than the first time interval needs to be preset. A sliding window approach is used, achieved by setting a fixed step size and window length. The window length can be set to the second or minute level, and the step size to half or less of the window length to ensure continuous data coverage. This sliding window approach dynamically acquires subsequences of each window within the second time interval, and then calculates a second statistical characteristic. This can include, but is not limited to, calculating statistical indicators such as standard deviation, variance, or root mean square as the second statistical characteristic to quantify the degree of data fluctuation within a short time window. The second statistical characteristic is then compared with a preset second threshold. This comparison can be achieved by calculating the deviation from the threshold in real time, or by combining multi-level threshold judgment logic to improve the sensitivity of anomaly detection. If the characteristic exceeds the threshold, a second-level warning signal is triggered. The core advantage of this mechanism lies in its high sensitivity to anomalies on short time scales. The continuous tracking of data fluctuations by the sliding window can effectively capture short-term anomalies such as transient shocks caused by mechanical loosening, avoiding sudden fluctuations that might be missed by fixed-window analysis.

[0032] This leads to the construction of a dual-channel parallel processing structure, enabling collaborative analysis of operational data across multiple time scales and fundamentally overcoming the blind spots in anomaly monitoring caused by traditional monitoring systems relying solely on a single time window.

[0033] The first data processing channel focuses on monitoring steady-state deviation anomalies over long time scales, enabling it to detect gradual changes such as bearing temperature rise and component wear. The second data processing channel focuses on monitoring transient impact anomalies over short time scales, enabling it to detect sudden events such as vibrations caused by mechanical loosening or electrical transient disturbances. This complementary time-scale structure significantly improves the monitoring capabilities for different types of anomalies without increasing hardware complexity. Furthermore, the integration of multi-scale analysis results through the joint criterion module further enhances the accuracy and reliability of fault diagnosis.

[0034] After obtaining the first-level and second-level warning signals, the equipment operating status category is determined based on the preset joint criterion module. The preset joint criterion module distinguishes between normal state, steady-state deviation anomaly, transient impact anomaly, or composite anomaly by combining the reception of the two warning signals through Boolean combination. In this way, the multi-scale analysis results are integrated, which can identify composite faults that cannot be captured by a single channel.

[0035] Finally, based on the determined equipment operating status category, control commands are generated according to the preset control strategy mapping table and sent to the equipment on the hydropower station side, ensuring that the control strategy is adapted to the anomaly type and effectively improving the pertinence of the response.

[0036] Therefore, this technical solution solves the problem that traditional monitoring systems cannot effectively monitor both long-term steady-state deviation anomalies and short-term transient impact anomalies of equipment simultaneously due to rigid data processing logic by using a dual-channel time-scale complementary design. It enables collaborative analysis of multi-time-scale operating data and command interaction.

[0037] Example 2 Based on Example 1, please refer to the following: Figure 2 The above is a flowchart illustrating the computer monitoring method for a hydropower centralized control center provided in an embodiment of the present invention. Further, the computer monitoring method for a hydropower centralized control center may specifically include the following contents.

[0038] Step S1: Obtain raw telemetry data from one or more devices deployed on the hydropower station side, configure timestamps, configure the corresponding monitoring type according to the preset device identification classification module, and obtain the raw data sequence; Step S2: In response to obtaining the original data sequence, the first data processing channel and the second data processing channel are started simultaneously; Step S3: The first data processing channel is configured to: preset a first time interval, obtain a first statistical feature quantity based on the original data sequence of any device within the first time interval; compare the first statistical feature quantity with a preset first threshold interval, and obtain a first-level warning signal if the threshold is exceeded. Step S4: The second data processing channel is configured to: preset a second time interval, obtain subsequences within each window through a sliding window method to obtain a second statistical feature; compare the second statistical feature with a preset second threshold, and if it exceeds the threshold, obtain a second-level warning signal; Step S5: Based on the first-level warning signal and the second-level warning signal, and based on the preset joint judgment module, determine the equipment operating status category, including one of the four categories: normal state, steady-state deviation anomaly, transient impact anomaly, or composite anomaly. Step S6: Based on the determined equipment operating status category and the preset control strategy mapping table, obtain control commands and send them to the equipment on the hydropower station side.

[0039] In this embodiment, it also includes: Step S7: In response to obtaining the original data sequence, the third data processing channel is started synchronously; The third data processing channel is configured as follows: Based on the monitoring type, a corresponding preset third time interval is obtained, and the length of any third time interval is within the length of the first time interval and the second time interval. Based on the original data sequence of any device within the third time interval, calculate the coefficient of variation to obtain the third statistical characteristic. When the third statistical feature is greater than the preset third threshold, it is determined that the device is in an abnormal operating state and the device is not in a normal operating state.

[0040] Step S8: Based on the original data sequence, obtain the real-time residual variance by training a time series model on historical data when the equipment operating status category is marked as normal. If the real-time residual variance exceeds the residual variance of the historical data when the equipment operating status category is marked as normal, and exceeds a set multiple in a continuous set sampling period, then the equipment operating status category is determined to be potentially abnormal, and a third-level warning signal is output.

[0041] Step S9: Update the first threshold interval and the second threshold, including: From the data of the past set time period when the equipment operating status category was marked as normal, all first statistical features and second statistical features are obtained to obtain a sample set; Based on the sample set, calculate the mean and standard deviation of the first statistical feature within a set time period to obtain the updated first threshold interval; Based on the sample set, calculate the second statistical feature within a set time period to obtain the updated second threshold. In response to the updated first threshold range and the updated second threshold being confirmed a second time by the operations and maintenance personnel, the current first threshold range and second threshold are obtained.

[0042] Based on the above steps, a mechanism is constructed to collaboratively analyze the operating status of equipment across multiple time scales and process the data, thereby enabling the monitoring method of dynamic evaluation of the operating status of equipment on the power station side and generation of control commands by the hydropower centralized control center computer.

[0043] This embodiment uses one or more mixed-flow turbine generator sets of a small hydropower station in a certain river basin as an example to illustrate this application with specific equipment.

[0044] In step S1, the raw telemetry data of one or more devices deployed on the hydropower station side are obtained, and the timestamps are configured. The corresponding monitoring type is configured according to the preset device identification classification module to obtain the raw data sequence. Raw telemetry data is received from multiple monitored devices, and data acquisition is performed at a fixed sampling period. The sampling frequency is set according to the monitored physical quantity. For example, for high-frequency signals such as unit vibration acceleration and bearing temperature, the sampling period is usually set to 10 to 100 milliseconds; while for low-frequency or quasi-static signals such as guide vane opening, power generation, and grid frequency, the sampling period can be relaxed to 500 milliseconds to 1 second.

[0045] After receiving all raw telemetry data, timestamps are configured, and a device ID is matched first through the device identification classification module. Then, the corresponding monitored physical quantity is found according to the device type, classified and marked, and then stored in the log storage repository to ensure efficient query and backtracking capabilities for historical data.

[0046] In step S2, in response to obtaining the original data sequence, the first data processing channel and the second data processing channel are started simultaneously; After data acquisition and preprocessing are completed, the first and second data processing channels are started simultaneously to perform long-term trend analysis and short-term transient feature extraction, respectively, and then the process jumps to steps S3 and S4.

[0047] In this embodiment, a third data processing channel can also be enabled simultaneously to enhance the ability to identify fluctuations at medium time scales, and then proceed to step S7 for execution.

[0048] In one possible implementation, in step S3, the first data processing channel is configured to: preset a first time interval; obtain a first statistical feature quantity based on the original data sequence of any device within the first time interval; compare the first statistical feature quantity with a preset first threshold interval; and obtain a first-level warning signal if the threshold is exceeded, including: Based on the monitoring type of the original data sequence, a first time interval is set; in this embodiment, the first time interval can be set to 900 seconds for bearing temperature monitoring, 600 seconds for guide vane opening monitoring, and 1800 seconds for power generation monitoring.

[0049] The raw data sequence within the first time interval is obtained from any device at a set sampling period, and the validity is filtered to obtain the valid data sequence; Based on the effective data sequence, the arithmetic mean of the effective data sequence is calculated to obtain the first statistical characteristic quantity, which characterizes the steady state of the equipment's operating state over a long period of time. This mainly refers to the steady state deviation caused by some normal losses during the operation of hydropower station equipment, reflecting steady state deviations such as slow temperature rise of bearings, parameter drift caused by component wear, or jamming of actuators. Based on the first statistical feature, a real-time comparison is performed based on a preset first threshold interval. The initial value of this interval needs to be determined based on the specific equipment, the equipment manufacturer's technical specifications, and historical normal operation data. If the first statistical feature is not within the preset first threshold range, it is determined that a steady-state shift anomaly has occurred. In response to a steady-state shift, a first-level warning signal is triggered.

[0050] Upon triggering the first-level warning signal, an automatic record is generated in the abnormal event log. This record includes at least the timestamp of the abnormality trigger, the associated device identifier, the type of abnormal physical quantity, the specific value of the first statistical characteristic quantity, and the magnitude of its deviation from the threshold range.

[0051] In the implementation of this application embodiment, a first time interval is set in the first data processing channel. Its length is determined based on the steady-state response characteristics of the monitored physical quantity. Taking the mixed-flow turbine generator set in this embodiment as an example, then for bearing temperature monitoring, It was set to 900 seconds (15 minutes); for guide vane opening monitoring, It is set to 600 seconds (10 minutes); for power generation monitoring, It was set to 1800 seconds (30 minutes).

[0052] For any device in the first time interval Raw data sequence collected internally Calculate its first statistical characteristic. That is, the arithmetic mean of this sequence, can be expressed as: ; Where n is the time interval The number of valid sampling points within the range. Valid sampling points are defined as data points that have not been marked as communication interruption, sensor failure, or data verification failure.

[0053] This calculation process can be executed using a dedicated data processing thread, employing a sliding window mechanism where the window advances forward by one sampling period at a time, thereby achieving... Continuous updates.

[0054] Then the real-time calculations obtained With the preset first threshold range This threshold range is determined during the initialization phase based on the technical specifications provided by the equipment manufacturer and historical data. For example, for the upper guide bearing of a certain type of water turbine, its normal operating temperature range is 40℃ to 70℃, then the initial... .like or If the first data processing channel outputs a first-level warning signal to the joint criterion module, it records the timestamp, device ID, physical quantity type, and deviation range in the abnormal event log.

[0055] Step S4 is a parallel transient anomaly monitoring process based on a short time scale. This process, together with the steady-state analysis process in step S3, constitutes a multi-time-scale collaborative analysis structure. In one possible implementation, in step S4, the second data processing channel is configured as follows: a second time interval is preset, and the second time interval is shorter than a preset first time interval. A subsequence within each window is obtained using a sliding window method to obtain a second statistical feature. Based on the second statistical feature, it is compared with a preset second threshold. If it exceeds the threshold, a second-level warning signal is obtained, including: Based on the first time interval, a second time interval shorter than the length of the first time interval is defined; Based on the original data sequence, the step size is set according to the length of the sampling period, and the second time interval is used as a sliding window to continuously move and obtain the continuous data subsequence within each window; Based on continuous data subsequences, the standard deviation of each continuous data subsequence is calculated to obtain a second statistical characteristic, which characterizes the severity of data fluctuations within a very short time window. An online recursive algorithm can then be used to reduce computational complexity and ensure real-time performance. This second statistical characteristic primarily addresses sudden data fluctuations with extremely short time windows caused by impact vibrations due to mechanical loosening, transient disturbances in electrical systems, or high-frequency spike noise generated by momentary sensor failures.

[0056] Based on the second statistical characteristic quantity and a preset second threshold, a real-time comparison is performed. The initial value of this threshold is set based on the historical fluctuation level of the device during steady-state operation. If the second statistical feature is not within the preset second threshold range, it is determined that there is a transient shock anomaly. In response to the presence of a transient impact anomaly, a level 2 warning signal is triggered.

[0057] In the implementation of this embodiment, simultaneously, a second time interval is set in the second data processing channel. Its length is significantly smaller than The value is typically between 1 and 10 seconds. In this embodiment, for vibration acceleration signals, Set to 2 seconds; for grid frequency signals, It was set to 5 seconds.

[0058] Second time interval The timeline is continuously moved using a sliding window, with a step size equal to the sampling period of the original data. For each subsequence at each sliding position... ; Where m is the number of sampling points within the window, for example, when the sampling period is 10ms, hour, First, calculate the local mean. , can be represented as: ; Then calculate the second statistical characteristic. In other words, the local standard deviation of this subsequence can be expressed as: ; The calculation process uses a recursive algorithm to reduce computational complexity, such as using the Welford iteration method to update the mean and variance in constant time.

[0059] Real-time calculation With the preset second threshold This threshold is set based on the historical fluctuation level of the equipment during steady-state operation, for example, the historical vibration acceleration signal of a unit under rated load. The mean is 0.15 m / s², and the standard deviation is 0.03 m / s². Therefore, the initial... It can be set to 0.25 m / s², which can be directly set to approximately the mean plus 3 times the standard deviation.

[0060] like Then the second data processing channel outputs a second-level early warning signal to the joint criterion module.

[0061] In one possible implementation, in step S5, based on the first-level warning signal and the second-level warning signal, and using a preset joint criterion module, the equipment operating status category is determined, including: The preset joint criterion module is configured as follows: Upon receiving a Level 1 warning signal, it is determined that the first statistical feature is not within the first threshold range; Upon receiving a second-level warning signal, it is determined that the second statistical feature is greater than the second threshold. By taking whether the first-level warning signal and the second-level warning signal are received as conditions, a Boolean combination is performed to obtain four equipment operating status categories, including normal state, steady-state deviation anomaly, transient impact anomaly, or composite anomaly. The first-level warning signal and the second-level warning signal are input into the preset joint criterion module, which outputs the equipment operation status category.

[0062] In one possible implementation, the reception of a first-level warning signal and a second-level warning signal are used as conditions for a Boolean combination to obtain four equipment operating state categories, including normal state, steady-state deviation anomaly, transient impact anomaly, or combined anomaly, including: If neither the first-level warning signal nor the second-level warning signal is received, the state is considered normal. If a Level 1 warning signal is received but a Level 2 warning signal is not received, the steady-state offset is determined to be abnormal. If no Level 1 warning signal is received, but a Level 2 warning signal is received, it is determined to be a transient impact anomaly. If a Level 1 warning signal is received and a Level 2 warning signal is received, it is determined to be a composite anomaly. The first-level warning signal and the second-level warning signal are input into the preset joint criterion module, which outputs the equipment operation status category. When the status category is output, a confidence score is also output. Based on the first statistical feature, the second statistical feature, the first threshold interval, and the second threshold, the degree to which the first statistical feature deviates from the first threshold interval and the extent to which the second statistical feature exceeds the second threshold are calculated to obtain a confidence score, which characterizes the accuracy of the determined state category.

[0063] In the implementation of this application embodiment, the joint criterion module receives statistical feature quantities and early warning signals from the first data processing channel and the second data processing channel, and determines the state category according to preset logical rules.

[0064] The state category determination unit executes the following determination process: First, it determines... Is it within the first threshold interval? Internally, simultaneously judge Is it less than or equal to the second threshold? Based on the Boolean combination of the two conditions, the device operating status is divided into four categories: Normal state and ; Steady-state offset anomaly: and ; Transient impact anomaly: and ; Composite anomaly: and .

[0065] The determination result is encapsulated into a structured status message, including status category, confidence score (calculated based on the degree to which the feature quantity deviates from the threshold), timestamp, and associated device information, and then stored in the log repository. Subsequent control command generation is also based on this status message.

[0066] After the status determination is completed, the control command generation generates corresponding control actions based on the control strategy mapping table. This mapping table is a pre-configured rule table, the contents of which are jointly defined by the hydropower station operation procedures and equipment protection logic. In step S6, based on the determined equipment operating status category and the preset control strategy mapping table, control commands are obtained and sent to the equipment on the hydropower station side.

[0067] In this embodiment, the control policy mapping table includes at least the following: When the state category is steady-state offset anomaly and the physical quantity is guide vane opening, the generated command is to fine-tune the guide vane servo valve opening by 1%; When the state category is transient shock anomaly and the physical quantity is vibration acceleration, the generated instruction is to reduce the load to 80% of the current level and start the online diagnostic program. When the status category is a composite anomaly, regardless of the physical quantity type, an instruction is generated to stop the machine immediately and lock the automatic restart function.

[0068] All generated control commands undergo a dual verification mechanism before being issued. This mechanism can perform two independent verifications: a security operation boundary verification and an identity authentication and integrity verification.

[0069] The safety operation boundary verification checks whether the command parameters meet the preset safety constraints. For example, the magnitude of the load adjustment command must not exceed ±10% per time; the guide vane opening adjustment rate must not exceed 2% per second; the emergency shutdown command is only allowed to be triggered when the unit load is higher than 30%; if the command violates any safety boundary, it will be rejected and recorded in the safety audit log.

[0070] Identity authentication and integrity verification perform cryptographic verification of the source and content of the instructions. All control instructions are digitally signed upon generation. This signature is generated by encrypting the SHA-256 hash value of the instruction content using the private key of the central control center. It is then transmitted to the controller of the downstream equipment on the power plant side through an encrypted communication channel. The corresponding public key is used to verify the validity of the signature and compare the instruction hash value to confirm that the content has not been tampered with.

[0071] The instruction is only allowed to be executed after both checks pass. The controller of the equipment at the lower level on the power station side performs the corresponding operation and feeds back the execution result to the central control center, forming a closed-loop control.

[0072] In one possible implementation, in step S7, in response to obtaining the original data sequence, the third data processing channel is synchronously started; The third data processing channel is configured as follows: Based on the monitoring type, a corresponding preset third time interval is obtained, and the length of any third time interval is within the length of the first time interval and the second time interval. Based on the original data sequence of any device within the third time interval, calculate the coefficient of variation to obtain the third statistical characteristic. When the third statistical feature is greater than the preset third threshold, it is determined that the device is in an abnormal operating state and the device is not in a normal operating state.

[0073] In the implementation of this application's embodiments, a third data processing channel is also introduced for calculating the local coefficient of variation at a medium time scale. .

[0074] Third time interval The length is between and The interval is typically between 30 and 120 seconds, for vibration signals. For temperature signals, .

[0075] Calculate the mean of the data series within this time window. with standard deviation And from this, we obtain the coefficient of variation, which can be expressed as: ; This index is dimensionless and suitable for comparing the stability across physical quantities. Compared with the preset third threshold Comparison, for example, setting This indicates that a relative fluctuation of 5% is allowed.

[0076] like If the value is not found, it indicates that the physical quantity has abnormal fluctuations on a medium time scale, which may indicate problems such as decreased lubrication efficiency or unstable cooling water flow. This information can be used as an auxiliary criterion input to the joint criterion module to improve the early identification capability of progressive faults.

[0077] In one possible implementation, in step S8, based on the original data sequence, a time series model trained on historical data when the equipment operating state category is marked as normal is obtained to obtain the real-time residual variance; the historical data when the equipment operating state category is marked as normal is the historical normal data, which is used to train the time series model.

[0078] The residual variance obtained from historical data when the equipment operating status category was marked as normal is the historical normal residual variance. When the real-time residual variance exceeds the historical normal residual variance and exceeds a set multiple within a continuously set sampling period, the equipment operating status is determined to be potentially abnormal, and a third-level warning signal is output.

[0079] In the implementation of this application embodiment, a threshold update is also introduced to overcome the problem that fixed thresholds may fail due to equipment aging or environmental changes during long-term operation.

[0080] Perform a threshold recalibration process periodically, for example, every 7 days, using the first threshold range. For example, extract all data from the data marked as normal within the past 30 days. Values ​​constitute the sample set Then calculate the sample mean. with sample standard deviation Then, based on the normal distribution, the new threshold is set as follows: This indicates that the normal fluctuation range is allowed to be covered at a preset confidence level.

[0081] Similarly, the second threshold Under normal historical conditions Update within the set values.

[0082] All threshold update operations must be confirmed a second time by operations and maintenance personnel, and the version change log must be recorded.

[0083] In one possible implementation, step S9, updating the first threshold interval and the second threshold, includes: From the data of the past set time period when the equipment operating status category was marked as normal, all first statistical features and second statistical features are obtained to obtain a sample set; Based on the sample set, calculate the mean and standard deviation of the first statistical feature within a set time period to obtain the updated first threshold interval; Based on the sample set, calculate the second statistical feature within a set time period to obtain the updated second threshold. In response to the updated first threshold range and the updated second threshold being confirmed a second time by the operations and maintenance personnel, the current first threshold range and second threshold are obtained.

[0084] In the implementation of this application, residual analysis is also introduced to monitor novel anomaly patterns that are not covered by the statistical model.

[0085] Residual analysis is based on, and is first trained, by applying, for any physical quantity Select a continuous data segment of no less than 7 days from historical normal data, and directly determine the order p and q of the autoregressive moving average model using the AIC criterion.

[0086] Taking a bearing temperature signal as an example, the optimal model, after determining p and q, is ARMA(2,1), which can be expressed as: ; Among them, coefficient The result was obtained by fitting the data using the maximum likelihood estimation method.

[0087] Then calculate the predicted value in real time. Compared with actual observed values residuals between And monitor the statistical properties of the residual sequence.

[0088] If the residual variance exceeds twice the historical normal residual variance within 10 consecutive sampling periods, or if there are 5 consecutive positive residuals, this indicates a systematic bias in terms of repeatability, unidirectionality, measurability, and so on. In this case, the residual analysis will output a potential abnormal signal to the joint criterion module, triggering a level 3 warning and prompting maintenance personnel to perform manual diagnosis.

[0089] Example 3 This is the third embodiment of the present invention. Based on embodiments 1 and 2, please refer to the following references. Figure 3 This embodiment provides a computer monitoring system for a hydropower centralized control center, including a data acquisition unit, a data processing unit, a first data processing channel unit, a second data processing channel unit, a unit for determining equipment operating status, and an instruction issuing unit that are connected in sequence by electricity.

[0090] The data acquisition unit is configured to: acquire raw telemetry data from one or more devices deployed on the side of the hydropower station, configure timestamps, configure the corresponding monitoring type according to the preset device identification classification module, and obtain the raw data sequence; The data processing unit is configured to: synchronously start the first data processing channel and the second data processing channel in response to obtaining the original data sequence; The first data processing channel unit is configured to: preset a first time interval, obtain a first statistical feature quantity based on the original data sequence of any device within the first time interval; compare the first statistical feature quantity with a preset first threshold interval, and obtain a first-level warning signal if the threshold is exceeded. The second data processing channel unit is configured to: preset a second time interval, obtain subsequences within each window through a sliding window method to obtain a second statistical feature quantity; compare the second statistical feature quantity with a preset second threshold, and obtain a second-level warning signal if the threshold is exceeded; The device operation status determination unit is configured to: determine the device operation status category based on the first-level warning signal and the second-level warning signal, and based on the preset joint criterion module; The instruction issuing unit is configured to: determine the equipment operating status category, obtain control instructions based on a preset control strategy mapping table, and issue them to the equipment on the hydropower station side.

[0091] In the implementation of this application embodiment, the data acquisition unit obtains raw telemetry data such as bearing temperature, vibration acceleration, guide vane opening, and power generation of each unit through a distributed sensor network. The data is then configured with millisecond-level high-precision timestamps and categorized by equipment ID and physical quantity type using an equipment identification and classification module, forming a structured raw data sequence. Specifically, the sampling period for high-frequency signals such as vibration acceleration and bearing temperature is set to 20ms, while the sampling period for low-frequency signals such as guide vane opening and power generation is set to 500ms.

[0092] After receiving the original data sequence, the data processing unit synchronously starts dual data processing channels.

[0093] The first data processing channel unit is preset with a first time interval. Bearing temperature monitoring was set to 900 seconds, guide vane opening monitoring to 600 seconds, and power generation monitoring to 1800 seconds. The arithmetic mean of the effective sampling points within each interval was calculated as the first statistical characteristic. Preset first threshold range For example, if the bearing temperature is 40℃-70℃, if If the distance exceeds this range, the first-level warning signal will be triggered.

[0094] The second data processing channel unit is preset with a second time interval. The vibration acceleration was set to 2 seconds, the power grid frequency to 5 seconds, and a sliding window with the sampling period as the step size was used to extract the subsequence of each window. The local mean was calculated first. Then, the local standard deviation is obtained using the variance formula and used as the second statistical characteristic. Preset second threshold ,like If the threshold is exceeded, a second-level warning signal will be triggered.

[0095] The unit for determining the operating status of the equipment is based on the joint criterion module. It determines four types of states according to the combination of dual warning signals: if neither signal is present, it is a normal state; if only the first-level warning is present, it is a steady-state offset anomaly; if only the second-level warning is present, it is a transient impact anomaly; and if both signals are present, it is a composite anomaly.

[0096] The instruction issuing unit generates instructions based on a preset control strategy mapping table: When the steady-state offset is abnormal and the guide vane opening is adjusted, the guide vane servo valve opening is increased by 1%. When a transient impact occurs and vibration acceleration is detected, the load is reduced to 80% and an online diagnostic program is initiated. In the event of a complex anomaly, an emergency shutdown is directly issued and the automatic restart function is locked. The command is sent to the controller of the lower-level power station-side equipment through an encrypted channel to achieve closed-loop control.

[0097] Through the above technical solutions, a computer monitoring method for hydropower centralized control centers was constructed that is compatible with new-generation computing platforms and has dynamic data processing capabilities, realizing collaborative analysis of multi-timescale operational data and safe and controllable command interaction.

[0098] Example 4 The fourth embodiment of the present invention differs from the previous embodiments in that: Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0099] In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices or units, or may be electrical, mechanical or other forms of connection.

[0100] The units described as separate components may or may not be physically separate. As will be apparent to those skilled in the art, the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0101] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0102] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a server, or a grid device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0103] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A computer monitoring method for a hydropower centralized control center, characterized in that, include: Acquire raw telemetry data from one or more devices deployed on the side of the hydropower station, configure timestamps, configure the corresponding monitoring type according to the preset device identification classification module, and obtain the raw data sequence; In response to obtaining the original data sequence, the first data processing channel and the second data processing channel are simultaneously started; The first data processing channel is configured to: preset a first time interval, obtain a first statistical feature quantity based on the original data sequence of any device within the first time interval; compare the first statistical feature quantity with a preset first threshold interval, and obtain a first-level warning signal if the threshold is exceeded. The second data processing channel is configured to: preset a second time interval, and obtain the subsequence within each window through a sliding window method to obtain the second statistical feature; Based on the second statistical feature, it is compared with a preset second threshold. If it exceeds the threshold, a second-level warning signal is obtained. Based on the first-level warning signal and the second-level warning signal, and using a preset joint criterion module, the equipment operating status category is determined; Based on the determined equipment operating status category, and using a preset control strategy mapping table, control commands are obtained and sent to the equipment on the hydropower station side.

2. The computer monitoring method for a hydropower centralized control center according to claim 1, characterized in that, It also includes synchronously starting a third data processing channel in response to obtaining the original data sequence; The third data processing channel is configured as follows: According to the monitoring type, a corresponding preset third time interval is obtained, and the length of any third time interval is within the length of the first time interval and the second time interval; Based on the original data sequence of any device within the third time interval, calculate the coefficient of variation to obtain the third statistical characteristic. When the third statistical feature is greater than the preset third threshold, it is determined that the device operating status is abnormal and the device operating status category is not normal.

3. The computer monitoring method for a hydropower centralized control center according to claim 1, characterized in that, The first data processing channel is configured to: preset a first time interval, and obtain a first statistical feature quantity based on the original data sequence of any device within the first time interval; Based on the first statistical feature, a comparison is made with a preset first threshold range. If the range is exceeded, a first-level warning signal is obtained, including: Based on the monitoring type of the original data sequence, a first time interval is set; The raw data sequence within the first time interval is obtained from any device at a set sampling period, and the validity is filtered to obtain the valid data sequence; Based on the effective data sequence, the arithmetic mean of the effective data sequence is calculated to obtain a first statistical characteristic quantity to characterize the steady state of the device's operating state over a long period of time; Based on the first statistical feature, a real-time comparison is performed within a preset first threshold range; If the first statistical feature is not within the preset first threshold range, it is determined that a steady-state shift anomaly has occurred. In response to a steady-state shift, a first-level warning signal is triggered.

4. The computer monitoring method for a hydropower centralized control center according to claim 1, characterized in that, The second data processing channel is configured to: preset a second time interval, which is smaller than the preset first time interval, and obtain the subsequence within each window by means of a sliding window to obtain the second statistical feature quantity; Based on the second statistical feature, a comparison is made with a preset second threshold. If the threshold is exceeded, a second-level warning signal is obtained, including: Based on the first time interval, a second time interval shorter than the length of the first time interval is defined; Based on the original data sequence, the step size is set according to the length of the sampling period, and the second time interval is used as a sliding window to move and obtain a continuous data subsequence within each window; Based on the continuous data subsequences, the standard deviation of each continuous data subsequence is calculated to obtain a second statistical characteristic quantity, which characterizes the degree of data fluctuation of the device within a very short time window; Based on the second statistical feature, a real-time comparison is performed using a preset second threshold. If the second statistical feature is not within the preset second threshold range, it is determined that there is a transient impact anomaly. In response to the presence of a transient impact anomaly, a level 2 warning signal is triggered.

5. The computer monitoring method for a hydropower centralized control center according to claim 1, characterized in that, The step of determining the equipment operating status category based on the first-level warning signal and the second-level warning signal, using a preset joint criterion module, includes: The preset joint criterion module is configured as follows: Upon receiving a Level 1 warning signal, it is determined that the first statistical feature is not within the first threshold range; Upon receiving a second-level warning signal, it is determined that the second statistical feature is greater than the second threshold. By taking whether the first-level warning signal and the second-level warning signal are received as conditions, a Boolean combination is performed to obtain four equipment operating status categories, including normal state, steady-state deviation anomaly, transient impact anomaly, or composite anomaly. The first-level warning signal and the second-level warning signal are input into the preset joint criterion module, which outputs the equipment operating status category.

6. The computer monitoring method for a hydropower centralized control center according to claim 5, characterized in that, The process involves using the receipt of a first-level warning signal and a second-level warning signal as conditions, performing a Boolean combination to obtain four equipment operating state categories, including normal state, steady-state deviation anomaly, transient impact anomaly, or combined anomaly, including: If neither the first-level warning signal nor the second-level warning signal is received, the state is considered normal. If a Level 1 warning signal is received but a Level 2 warning signal is not received, the steady-state offset is determined to be abnormal. If no Level 1 warning signal is received, but a Level 2 warning signal is received, it is determined to be a transient impact anomaly. If a Level 1 warning signal is received and a Level 2 warning signal is received, it is determined to be a composite anomaly. The first-level warning signal and the second-level warning signal are input into the preset joint criterion module, which outputs the equipment operation status category. When the status category is output, a confidence score is also output. Based on the first statistical feature, the second statistical feature, the first threshold interval, and the second threshold, the degree to which the first statistical feature deviates from the first threshold interval and the extent to which the second statistical feature exceeds the second threshold are calculated to obtain a confidence score, which characterizes the accuracy of the determined state category.

7. The computer monitoring method for a hydropower centralized control center according to claim 1, characterized in that, It also includes updating the first threshold interval and the second threshold, including: From the data of the past set time period when the equipment operating status category was marked as normal, all first statistical features and second statistical features are obtained to obtain a sample set; Based on the sample set, calculate the mean and standard deviation of the first statistical feature within the set time period to obtain the updated first threshold interval; Based on the sample set, calculate the second statistical feature quantity within the set time period to obtain the updated second threshold; In response to the updated first threshold range and the updated second threshold being confirmed a second time by the operations and maintenance personnel, the current first threshold range and second threshold are obtained.

8. The computer monitoring method for a hydropower centralized control center according to claim 1, characterized in that, Also includes: Based on the original data sequence, a time series model trained on data when the equipment operating status category is marked as normal is used to obtain the real-time residual variance. If the real-time residual variance exceeds the residual variance of the historical data when the device operating status category is marked as normal, and exceeds a set multiple for a continuous set sampling period, then the device operating status category is determined to be potentially abnormal, and a third-level warning signal is output.

9. A computer monitoring system for a hydropower centralized control center, characterized in that, It includes a data acquisition unit, a data processing unit, a first data processing channel unit, a second data processing channel unit, a device operating status determination unit, and an instruction issuing unit, all connected in sequence by electrical connection. The data acquisition unit is configured to: acquire raw telemetry data from one or more devices deployed on the side of the hydropower station, configure timestamps, configure corresponding monitoring types according to the preset device identification classification module, and obtain raw data sequences; The data processing unit is configured to: in response to obtaining the original data sequence, simultaneously start the first data processing channel and the second data processing channel; The first data processing channel unit is configured to: preset a first time interval, obtain a first statistical feature quantity based on the original data sequence of any device within the first time interval; compare the first statistical feature quantity with a preset first threshold interval, and obtain a first-level warning signal if the threshold is exceeded; The second data processing channel unit is configured to: preset a second time interval, and obtain the subsequence within each window through a sliding window to obtain a second statistical feature; Based on the second statistical feature, it is compared with a preset second threshold. If it exceeds the threshold, a second-level warning signal is obtained. The device operation status determination unit is configured to: determine the device operation status category based on the first-level warning signal and the second-level warning signal, and based on a preset joint criterion module; The instruction issuing unit is configured to: obtain control instructions based on the determined equipment operating status category and a preset control strategy mapping table, and issue them to the equipment on the hydropower station side.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 8.

Images