Unit operating status monitoring method based on multi-dimensional model prediction

Abstract

The present invention discloses a method for monitoring the operating status of a unit based on multidimensional model prediction; it relates to the field of multidimensional model early warning of thermal power units; the present invention obtains key equipment of the unit and monitoring parameters of the key equipment, sets monitoring points corresponding to the monitoring parameters, and then obtains the equipment operation network of the unit; sets parameter association groups and training cycles of the equipment operation network, and obtains parameter data corresponding to the parameter association groups according to the training cycles; obtains association functions of the parameter association groups according to the parameter data, and then obtains the multidimensional model of the unit; monitors the operating status of the unit based on the multidimensional model to obtain abnormal operating status; generates abnormal early warning signals according to the abnormal operating status; the present invention effectively improves the comprehensiveness of the monitoring of the operating status of the thermal power unit.

Description

Technical Field

[0001] The present invention relates to the field of multi-dimensional model early warning for thermal power units, and in particular to a method for monitoring the operating status of a unit based on multi-dimensional model prediction. Background Art

[0002] In large thermal power plants, the safe, stable, and efficient operation of the unit (typically referring to the system consisting of boilers, turbines, generators, and their auxiliary components) is crucial. Currently, condition monitoring of thermal power units primarily relies on distributed control systems (DCSs) that provide single-point parameter threshold alarms (e.g., temperature, pressure, vibration, and flow rate violations) for key equipment (e.g., boilers, turbines, feedwater pumps, and fans). However, this approach has significant limitations:

[0003] Isolated alarms: Alarms are often triggered based on the instantaneous value of a single measurement point exceeding a static threshold, ignoring the complex physical connections between devices and the inherent dynamic coupling between operating parameters. This results in isolated and fragmented alarm information, making it difficult to accurately determine the root cause of the fault and predict its development trend.

[0004] Lag and passivity: Single-point threshold alarms are often triggered only after equipment failure has already occurred or operational conditions have significantly deteriorated. This creates a short warning window, leaving operators with limited time to intervene and make adjustments, making it difficult to achieve true predictive maintenance and proactive prevention. For example, in the early stages of a micro-leak on a boiler's heating surface, it may manifest as only a slight abnormality in localized wall temperature or subtle changes in related parameters. Traditional methods struggle to effectively capture these early signs.

[0005] False alarms and missed alarms: Static thresholds are difficult to adapt to the normal operation fluctuations of the unit under different loads and operating conditions (start-stop, variable load), and are prone to false alarms. At the same time, traditional methods are prone to missed alarms for potential faults caused by slight deviations of multiple parameters or destruction of correlation relationships.

[0006] Therefore, the thermal power industry urgently needs an intelligent monitoring method that can transcend the limitations of single-point parameters and deeply explore the correlations between equipment and the dynamic coupling patterns between parameters. This method should be able to learn the "health status" model of the unit based on historical operating data, evaluate the operational coordination and parameter matching of the entire system in real time, and identify abnormal patterns and potential risks before or at the earliest stages of a fault. This will enable earlier warning, more accurate positioning, and more proactive decision-making, effectively improving the operational safety, reliability, and economic efficiency of thermal power units. Summary of the Invention

[0007] In order to solve the above technical problems, the present invention provides a unit operation status monitoring method based on multi-dimensional model prediction;

[0008] The object of the present invention can be achieved by the following technical solution: a method for monitoring the operating status of a unit based on multi-dimensional model prediction, the method comprising the following steps:

[0009] Step S1: Acquire key equipment of the unit, obtain monitoring parameters of the key equipment, set monitoring points corresponding to the monitoring parameters, and then obtain the equipment operation network of the unit;

[0010] Step S2: Setting the parameter association group and training cycle of the equipment operation network, obtaining parameter data corresponding to the parameter association group according to the training cycle; obtaining the correlation function of the parameter association group according to the parameter data, and then obtaining the multi-dimensional model of the unit;

[0011] Step S3: monitoring the unit operating status based on the multi-dimensional model to obtain abnormal operating status;

[0012] Step S4: Generate an abnormal warning signal according to the abnormal operating state.

[0013] Furthermore, the process of obtaining the key equipment of the unit includes:

[0014] Set an abnormal frequency monitoring cycle, obtain the historical monitoring records of the unit based on the abnormal frequency monitoring cycle, and then obtain the abnormal frequency of the corresponding equipment of each unit based on the historical monitoring records; set the abnormal impact degree and abnormal maintenance cost of each equipment; and then obtain the abnormal cost value of each equipment by multiplying the abnormal frequency, abnormal impact degree and abnormal maintenance cost;

[0015] Set an abnormal cost value threshold. If the abnormal cost value exceeds the abnormal cost value threshold, the corresponding device will be marked as a key device; otherwise, it will be marked as a secondary device.

[0016] Furthermore, the process of setting monitoring points corresponding to monitoring parameters and obtaining the equipment operation network of the unit includes:

[0017] Generate corresponding device nodes for key equipment and secondary equipment, mark them as key equipment nodes and secondary equipment nodes respectively, and associate the key equipment nodes and secondary equipment nodes according to the unit's operating mechanism to generate a series device network and a synchronous device network;

[0018] The monitoring points of the corresponding monitoring parameters of each device are connected to the corresponding device nodes to generate the device operation network.

[0019] Furthermore, the process of setting the parameter association group of the device operation network includes:

[0020] The parameter association group includes a direct causal association group, an operation association group and a time sequence association group;

[0021] Based on the equipment operation network, the monitoring parameters corresponding to each key device are set as abnormal variable factors of the direct causal association group; the monitoring parameters associated with the abnormal factors in each secondary device are set as abnormal dependent variable factors, and then the abnormal variable factors and the abnormal dependent variable factors are integrated to generate a direct causal association group;

[0022] The same monitoring parameters of the synchronous device network are integrated to generate corresponding timing association groups; and all monitoring parameters corresponding to the first device of the serial device network are respectively associated data heads according to the serial connection sequence of the serial device network, and then the associated data heads are associated with the monitoring parameters of each device in the serial connection sequence to obtain the operation association group.

[0023] Furthermore, the process of obtaining parameter data corresponding to the parameter association group according to the training cycle includes:

[0024] According to the training cycle, several synchronous collection time points are set in the synchronous device network, and the training cycle is used as the total operating time of the serial device network. Then, the serial collection time points of each device in the serial device network are set. According to the synchronous collection time points, the parameter data corresponding to the timing association group and the direct causal association group are collected. According to the serial collection time points, the parameter data corresponding to the operation association group are collected.

[0025] Furthermore, the process of obtaining the correlation function of the parameter correlation group according to the parameter data and then obtaining the multi-dimensional model of the unit includes:

[0026] Obtain the ratio of the abnormal dependent variable factor to the abnormal variable factor in each direct causal association group based on the parameter data to generate the variable coefficient of the corresponding direct causal association group; schedule the same direct causal association group, and then obtain the average value of the variable coefficient and record it as the correlation coefficient; generate an abnormal linear function from the abnormal variable factor and the abnormal dependent variable factor based on the correlation variable coefficient; then schedule the abnormal linear function corresponding to the same abnormal dependent variable factor, and integrate them to generate the direct causal association function of the direct causal association group corresponding to the abnormal dependent variable factor;

[0027] Set the equalization parameter data threshold corresponding to the timing correlation group, calculate the difference between the parameter data of each monitoring parameter of the same timing correlation group and the corresponding equalization parameter data threshold to obtain the equalization difference; obtain the average value of each equalization difference of the timing correlation group as the timing correlation coefficient, and then obtain the timing correlation function corresponding to the timing correlation group based on the timing correlation coefficient;

[0028] Set the parameter data threshold range of the parameter data corresponding to each monitoring parameter of the operation association group, and compare the parameter data corresponding to each monitoring parameter of the operation association group with the corresponding parameter data threshold range in turn;

[0029] If the corresponding parameter data threshold range exists in sequence from the associated data header, the corresponding running association group is generated into a running association group function, which is recorded as 1;

[0030] If there is no corresponding parameter data threshold range starting from the association data header, the corresponding running association group is generated as a running association group function and recorded as 0;

[0031] The direct causal correlation function, time series correlation function and operation correlation group function are integrated to generate a multidimensional model.

[0032] Furthermore, the process of monitoring the unit operating status based on the multi-dimensional model to obtain abnormal operating status includes:

[0033] Set the abnormal threshold range corresponding to the abnormal dependent variable factor. If the abnormal dependent variable factor does not fall within the abnormal threshold range, the unit operation status is marked as abnormal operation status; otherwise, no action is taken;

[0034] Set the threshold range of each monitoring parameter of the time series correlation function. If the monitoring parameter does not fall within the threshold range, the unit operation status will be marked as abnormal; otherwise, no action will be taken.

[0035] If the operation association group function is 0, the unit operation status is marked as an abnormal operation status; otherwise, no processing is performed.

[0036] Furthermore, the process of generating an abnormal warning signal according to the abnormal operating state includes:

[0037] Send abnormal warning signals to the preset management terminal for management.

[0038] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention obtains the key equipment of the unit, obtains the monitoring parameters of the key equipment, sets the monitoring points corresponding to the monitoring parameters, and then obtains the equipment operation network of the unit; sets the parameter association group and training cycle of the equipment operation network, and obtains the parameter data corresponding to the parameter association group according to the training cycle; obtains the correlation function of the parameter association group according to the parameter data, and then obtains the multidimensional model of the unit; monitors the operating status of the unit based on the multidimensional model to obtain the abnormal operating status; generates an abnormal warning signal according to the abnormal operating status; the present invention effectively improves the comprehensiveness of the operating status monitoring of the thermal power unit and the accuracy of the monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction to the drawings required for use in the embodiments will be given below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0040] Figure 1 This is a flow chart of the unit operation status monitoring method based on multi-dimensional model prediction. DETAILED DESCRIPTION

[0041] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0042] like Figure 1 As shown, a method for monitoring the unit operating status based on multi-dimensional model prediction includes the following steps:

[0043] Step S1: Acquire key equipment of the unit, obtain monitoring parameters of the key equipment, set monitoring points corresponding to the monitoring parameters, and then obtain the equipment operation network of the unit;

[0044] Step S2: Setting the parameter association group and training cycle of the equipment operation network, obtaining parameter data corresponding to the parameter association group according to the training cycle; obtaining the correlation function of the parameter association group according to the parameter data, and then obtaining the multi-dimensional model of the unit;

[0045] Step S3: monitoring the unit operating status based on the multi-dimensional model to obtain abnormal operating status;

[0046] Step S4: Generate an abnormal warning signal according to the abnormal operating state.

[0047] Step S1 needs to be further defined. The process of obtaining the key equipment of the unit, obtaining the monitoring parameters of the key equipment, setting the monitoring points corresponding to the monitoring parameters, and then obtaining the monitoring data acquisition model of the unit includes:

[0048] Set an abnormal frequency monitoring cycle, obtain the historical monitoring records of the unit based on the abnormal frequency monitoring cycle, and then obtain the abnormal frequency of the corresponding equipment of each unit based on the historical monitoring records; set the abnormal impact degree and abnormal maintenance cost of each equipment; and then obtain the abnormal cost value of each equipment by multiplying the abnormal frequency, abnormal impact degree and abnormal maintenance cost;

[0049] Set an abnormal cost value threshold. If the abnormal cost value exceeds the abnormal cost value threshold, the corresponding device will be marked as a critical device; otherwise, it will be marked as a secondary device.

[0050] Generate corresponding device nodes for key equipment and secondary equipment, mark them as key equipment nodes and secondary equipment nodes respectively, and associate the key equipment nodes and secondary equipment nodes with each other according to the unit's operating mechanism to generate a series device network and a synchronous device network; the operating mechanism includes series device operation and synchronous device operation;

[0051] In the above embodiment, it should be further explained that the device nodes are associated to generate a series device network based on the operation of the series devices. For example, in a thermal power unit, the boiler, steam turbine, and generator must form a complete energy conversion chain (fuel combustion → steam drives the steam turbine → generator generates electricity). The three must operate simultaneously. The shutdown of any device will cause the entire unit to be paralyzed.

[0052] The monitoring points of the corresponding monitoring parameters of each device are connected to the corresponding device nodes to generate the device operation network.

[0053] In the above embodiment, it needs to be further explained that the equipment of the unit is connected through the operation mechanism to prepare for the subsequent monitoring of the operation status.

[0054] Step S2 needs to be further defined. The process of setting the parameter association group and training cycle of the equipment operation network, obtaining parameter data corresponding to the parameter association group according to the training cycle, and obtaining the correlation function of the parameter association group according to the parameter data, and then obtaining the multi-dimensional model of the unit includes:

[0055] The parameter association group includes a direct causal association group, an operation association group and a time sequence association group;

[0056] Based on the equipment operation network, the monitoring parameters corresponding to each key device are set as abnormal variable factors of the direct causal association group, denoted as KLi; where L represents the number of the device node corresponding to the key device; i represents the number of the monitoring point corresponding to the monitoring parameter in the key device, and both L and i are positive integers; then, the monitoring parameters associated with the abnormal factors in each secondary device are set as abnormal dependent variable factors, denoted as RSm KLi ; Wherein, S represents the number of the device node corresponding to the secondary device, m represents the number of the monitoring point corresponding to the monitoring parameter in the secondary device, and both S and m are positive integers; then, the abnormal variable factors and the abnormal dependent variable factors are integrated to generate a direct causal association group;

[0057] Integrate the same monitoring parameters of the synchronous device network to generate a corresponding time sequence association group; and according to the series connection sequence of the series device network, all monitoring parameters corresponding to the first device in the series device network are respectively associated data headers, and then the associated data headers are associated with the monitoring parameters of each device in the series connection sequence to obtain an operation association group;

[0058] In the above embodiment, it needs to be further explained that the timing association group is used to represent the parameter data of the same monitoring parameter in the operation of different devices when each device in the synchronous device network is running simultaneously. The association and changes of the monitoring parameters in the synchronous device network can be monitored through the parameter data; similarly, the functional association group is used to represent the serial device network. Since the operation of the serial device network is strictly based on the sequence of the devices, the monitoring parameters of the first device may cause changes in the subsequent monitoring parameters.

[0059] According to the training cycle, several synchronization acquisition time points are set in the synchronous device network, denoted as tj, where j = 1, 2, ..., h, and h is a positive integer; and the training cycle is used as the total operation time of the series device network, and then the series acquisition time point of each device in the series device network is set, denoted as td, where d represents the device number of the series device network, d = 1, 2, ..., r, and r is a positive integer;

[0060] Collect parameter data corresponding to the time series association group and the direct causal association group according to the synchronous collection time point, and collect parameter data corresponding to the operation association group according to the serial collection time point;

[0061] In the above embodiment, it needs to be further explained that different collection nodes need to be set for collection of data from different device networks during the training cycle, and then different parameter association groups are monitored to obtain corresponding parameter data.

[0062] According to the parameter data, the ratio of the abnormal dependent variable factor to the abnormal variable factor in each direct causal association group is obtained to generate the variable coefficient of the corresponding direct causal association group; the same direct causal association group is scheduled, and the average value of the variable coefficient is obtained and recorded as the correlation coefficient, which is recorded as a KLi ; According to the correlation variable coefficient, the abnormal variable factor and the abnormal dependent variable factor generate an abnormal linear function, which is recorded as QSm KLi =a KLi ×KLi; then dispatch the abnormal linear function corresponding to the same abnormal dependent variable factor, and integrate it to generate the direct causal correlation function of the direct causal correlation group corresponding to the abnormal dependent variable factor, recorded as ;

[0063] In the above embodiment, it needs to be further explained that the reason for using the monitoring parameters of key equipment as abnormal variable factors is that key equipment has a higher abnormal cost and a higher probability of abnormality. Therefore, using it as an abnormal variable factor further improves the abnormal correlation.

[0064] Set the equalization parameter data threshold corresponding to the timing association group, calculate the difference between the parameter data of each monitoring parameter of the same timing association group and the corresponding equalization parameter data threshold to obtain the equalization difference; obtain the average value of each equalization difference of the timing association group and record it as the timing association coefficient, and then obtain the timing association function corresponding to the timing association group according to the timing association coefficient, which is recorded as Among them, Q u It is represented as the u-th monitoring parameter Q in the timing association group; and u=1, 2, ..., v, where v is a positive integer; It is expressed as the time series correlation coefficient corresponding to the u-th monitoring parameter;

[0065] Set the parameter data threshold range of the parameter data corresponding to each monitoring parameter of the operation association group, and compare the parameter data corresponding to each monitoring parameter of the operation association group with the corresponding parameter data threshold range in turn;

[0066] If the corresponding parameter data threshold range exists in sequence from the associated data header, the corresponding running association group is generated into a running association group function, which is recorded as 1;

[0067] If there is no corresponding parameter data threshold range starting from the association data header, the corresponding running association group is generated as a running association group function and recorded as 0;

[0068] Integrate direct causal correlation function, temporal correlation function and operational correlation group function to generate a multidimensional model;

[0069] In the above embodiment, it needs to be further explained that the monitoring parameters corresponding to the equipment of the unit are integrated and monitored through different correlation functions to improve the monitoring capability.

[0070] Step S3 needs to be further defined. The process of monitoring the unit operating status based on the multi-dimensional model to obtain abnormal operating status includes:

[0071] Set the abnormal threshold range corresponding to the abnormal dependent variable factor. If the abnormal dependent variable factor does not fall within the abnormal threshold range, the unit operation status is marked as abnormal operation status; otherwise, no action is taken;

[0072] Set the threshold range of each monitoring parameter of the time series correlation function. If the monitoring parameter does not fall within the threshold range, the unit operation status will be marked as abnormal; otherwise, no action will be taken.

[0073] If the operation association group function is 0, the unit operation status is marked as an abnormal operation status; otherwise, no processing is performed.

[0074] Step S4 needs to be further defined. The process of generating an abnormal warning signal according to an abnormal operating state includes:

[0075] Send abnormal warning signals to the preset management terminal for management.

[0076] Working principle: The present invention obtains the key equipment of the unit and the monitoring parameters of the key equipment, sets the monitoring points corresponding to the monitoring parameters, and then obtains the equipment operation network of the unit; sets the parameter association group and training cycle of the equipment operation network, and obtains the parameter data corresponding to the parameter association group according to the training cycle; obtains the correlation function of the parameter association group according to the parameter data, and then obtains the multidimensional model of the unit; monitors the operating status of the unit based on the multidimensional model to obtain the abnormal operating status; and generates an abnormal warning signal according to the abnormal operating status.

[0077] The features and exemplary embodiments of various aspects of the present application are described in detail above. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is further described in detail in combination with the drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without some of these specific details. The above description of the embodiments is merely to provide a better understanding of the present application by showing examples of the present application.

[0078] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A unit operation status monitoring method based on multi-dimensional model prediction, characterized in that: The method comprises the following steps: Step S1: Acquire key equipment of the unit, obtain monitoring parameters of the key equipment, set monitoring points corresponding to the monitoring parameters, and then obtain the equipment operation network of the unit; Step S2: Setting the parameter association group and training cycle of the equipment operation network, obtaining parameter data corresponding to the parameter association group according to the training cycle; obtaining the correlation function of the parameter association group according to the parameter data, and then obtaining the multi-dimensional model of the unit; Step S3: monitoring the unit operating status based on the multi-dimensional model to obtain abnormal operating status; Step S4: generating an abnormal warning signal according to the abnormal operating state; The process of acquiring the key equipment of the unit includes: Set an abnormal frequency monitoring cycle, obtain the historical monitoring records of the unit based on the abnormal frequency monitoring cycle, and then obtain the abnormal frequency of the corresponding equipment of each unit based on the historical monitoring records; set the abnormal impact degree and abnormal maintenance cost of each equipment; and then obtain the abnormal cost value of each equipment by multiplying the abnormal frequency, abnormal impact degree and abnormal maintenance cost; Set an abnormal cost value threshold. If the abnormal cost value exceeds the abnormal cost value threshold, the corresponding device will be marked as a critical device; otherwise, it will be marked as a secondary device. The process of setting up parameter association groups for the device operation network includes: The parameter association group includes a direct causal association group, an operation association group and a time sequence association group; Based on the equipment operation network, the monitoring parameters corresponding to each key device are set as abnormal variable factors of the direct causal association group; and the monitoring parameters associated with the abnormal variable factors in each secondary device are set as abnormal dependent variable factors, and then the abnormal variable factors and the abnormal dependent variable factors are integrated to generate a direct causal association group; The same monitoring parameters of the synchronous device network are integrated to generate corresponding timing association groups; and all monitoring parameters corresponding to the first device of the serial device network are respectively associated data heads according to the serial connection sequence of the serial device network, and then the associated data heads are associated with the monitoring parameters of each device in the serial connection sequence to obtain the operation association group.

2. The method for monitoring the unit operating status based on multi-dimensional model prediction according to claim 1, characterized in that: The process of setting monitoring points corresponding to monitoring parameters and then obtaining the equipment operation network of the unit includes: Generate corresponding device nodes for key equipment and secondary equipment, mark them as key equipment nodes and secondary equipment nodes respectively, and associate the key equipment nodes and secondary equipment nodes according to the unit's operating mechanism to generate a series device network and a synchronous device network; The monitoring points of the corresponding monitoring parameters of each device are connected to the corresponding device nodes to generate the device operation network.

3. The method for monitoring the unit operating status based on multi-dimensional model prediction according to claim 1, characterized in that: The process of obtaining parameter data corresponding to the parameter association group according to the training cycle includes: According to the training cycle, several synchronous collection time points are set in the synchronous device network, and the training cycle is used as the total operating time of the serial device network. Then, the serial collection time points of each device in the serial device network are set. According to the synchronous collection time points, the parameter data corresponding to the timing association group and the direct causal association group are collected. According to the serial collection time points, the parameter data corresponding to the operation association group are collected.

4. The method for monitoring the unit operating status based on multi-dimensional model prediction according to claim 1, characterized in that: The process of obtaining the correlation function of the parameter correlation group based on the parameter data and then obtaining the multi-dimensional model of the unit includes: Obtain the ratio of the abnormal dependent variable factor to the abnormal variable factor in each direct causal association group based on the parameter data to generate the variable coefficient of the corresponding direct causal association group; schedule the same direct causal association group, and then obtain the average value of the variable coefficient and record it as the correlation coefficient; generate an abnormal linear function from the abnormal variable factor and the abnormal dependent variable factor based on the correlation variable coefficient; then schedule the abnormal linear function corresponding to the same abnormal dependent variable factor, and integrate them to generate the direct causal association function of the direct causal association group corresponding to the abnormal dependent variable factor; Set the equalization parameter data threshold corresponding to the timing correlation group, calculate the difference between the parameter data of each monitoring parameter of the same timing correlation group and the corresponding equalization parameter data threshold to obtain the equalization difference; obtain the average value of each equalization difference of the timing correlation group as the timing correlation coefficient, and then obtain the timing correlation function corresponding to the timing correlation group based on the timing correlation coefficient; Set the parameter data threshold range of the parameter data corresponding to each monitoring parameter of the operation association group, and compare the parameter data corresponding to each monitoring parameter of the operation association group with the corresponding parameter data threshold range in turn; If the corresponding parameter data threshold range exists in sequence from the associated data header, the corresponding running association group is generated into a running association group function, which is recorded as 1; If there is no corresponding parameter data threshold range starting from the association data header, the corresponding running association group is generated as a running association group function and recorded as 0; The direct causal correlation function, time series correlation function and operation correlation group function are integrated to generate a multidimensional model.

5. The method for monitoring the unit operating status based on multi-dimensional model prediction according to claim 1, characterized in that: The process of monitoring the unit operating status and obtaining abnormal operating status based on the multi-dimensional model includes: Set the abnormal threshold range corresponding to the abnormal dependent variable factor. If the abnormal dependent variable factor does not fall within the abnormal threshold range, the unit operation status is marked as abnormal operation status; otherwise, no action is taken; Set the threshold range of each monitoring parameter of the time series correlation function. If the monitoring parameter does not fall within the threshold range, the unit operation status will be marked as abnormal; otherwise, no action will be taken. If the operation association group function is 0, the unit operation status is marked as an abnormal operation status; otherwise, no processing is performed.

6. The method for monitoring the unit operating status based on multi-dimensional model prediction according to claim 1, characterized in that: The process of generating abnormal warning signals based on abnormal operating status includes: Send abnormal warning signals to the preset management terminal for management.

Images