Redundant control system of coal feeder based on DCS and SCM

Abstract

The present invention discloses a redundant control system for a coal feeder based on DCS and SCM, which relates to the field of coal machine control technology, including a standby unit self-check and status monitoring module, a main control unit real-time data acquisition module, a data preprocessing and quality improvement module, a key feature extraction and analysis module, a deep learning model prediction module, an operation status classification and decision module, a routine maintenance and monitoring module, and an abnormal operation self-check cycle adjustment module: the standby unit self-check and status monitoring module, first, executes a preset self-check and status monitoring cycle, and performs regular inspections on the standby unit to ensure that the standby unit is in normal working condition when on standby. The present invention uses regular self-checks and status monitoring, combined with deep learning to predict the abnormality of the main control unit, and dynamically adjusts the self-check frequency of the standby unit to ensure that it can reliably take over the task, improve system stability, reduce failures and production interruptions, and optimize economic benefits.

Description

Technical Field

[0001] The present invention relates to the technical field of coal machine control, and in particular to a redundant control system for a coal feeder based on DCS and SCM. Background Art

[0002] Coal feeder redundant control based on DCS and SCM refers to a redundant control scheme built using distributed control system (DCS) and single-chip microcomputer (SCM) technology to improve the reliability and stability of the coal feeder. In this system, DCS is used to manage and monitor the various control modules of the coal feeder, and real-time data acquisition, processing and feedback are performed through distributed control nodes to ensure the coordination and safety of the entire process. The single-chip microcomputer (SCM), as a low-cost and efficient microcontroller, is used in the local control link for precise operation control. The core idea of ​​redundant control is that when a certain control link fails, the redundant control unit (whether it is a DCS system or a single-chip microcomputer) can quickly take over to ensure the continuous operation of the coal feeder and avoid production stoppage or equipment damage caused by single-point failure. Through this redundant design, the reliability of the coal feeder system can be greatly improved, the risk of failure can be reduced, and the stable operation of the coal transportation system can be ensured.

[0003] The role of the backup unit is to ensure that the system can quickly and seamlessly take over the control task when the main control unit fails, so as to maintain the continuous operation of the coal feeder and avoid production interruption. DCS (distributed control system) is the main control platform responsible for monitoring and coordinating multiple control nodes, while SCM (single-chip microcomputer) is usually used as an auxiliary unit for local control. In the redundant control scheme, the backup unit can be another DCS control node or a backup SCM unit with the same function as the main system. The role of the backup unit is to quickly take over the main control task when the main unit fails, maintain the normal operation of the entire system through the redundant switching mechanism, and ensure the stability of coal supply. Under normal circumstances, the backup unit is on standby, but it always keeps synchronization with the main system and performs self-test and status monitoring regularly. Once the main control unit fails or an abnormality occurs, the backup unit will take over immediately to ensure that the operation of the coal feeder is not interrupted, reduce the impact of system failures on the production process, and avoid equipment damage or production stagnation caused by control failure.

[0004] The prior art has the following deficiencies:

[0005] The existing technology usually maintains the backup unit through regular self-checking and status monitoring to ensure that the backup unit can quickly and reliably take over the control task when the main control unit fails. However, when the main control unit fails and needs to switch to the backup unit, if the backup unit fails to perform checks in time based on the preset self-checking and status monitoring cycle, the backup unit may not be able to detect its hardware or software failures (such as processor failure, power supply problems, memory corruption or control program errors) in time. In this case, the backup unit cannot take over the control task normally, resulting in the failure of redundant protection and the system cannot resume normal operation. The coal supply will be interrupted, the coal feeder will stop working, and the entire production line will come to a complete standstill. The production process will suffer a long interruption, and the shutdown may last for hours to days, causing serious production losses and having a profound impact on equipment safety, system stability and economic benefits.

[0006] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0007] The purpose of the present invention is to provide a redundant control system for coal feeders based on DCS and SCM, which ensures that the standby unit works normally in the standby state by regularly performing self-inspection and status monitoring of the standby unit, while monitoring the operating status of the main control unit in real time, and combining the deep learning model to predict the abnormal performance of the main control unit, and dynamically adjust the self-inspection frequency of the standby unit. When the main control unit is abnormal, the self-inspection frequency of the standby unit is quickly increased to ensure that it can reliably take over the task, avoid the failure of redundant switching, and reduce the risk of production interruption and failure. This intelligent maintenance mechanism improves the reliability of the standby unit, saves maintenance costs, and optimizes system stability and economic benefits to solve the problems in the above-mentioned background technology.

[0008] In order to achieve the above-mentioned purpose, the present invention provides the following technical solutions: a redundant control system for a coal feeder based on DCS and SCM, including a standby unit self-check and status monitoring module, a main control unit real-time data acquisition module, a data preprocessing and quality improvement module, a key feature extraction and analysis module, a deep learning model prediction module, an operation status classification and decision module, a routine maintenance and monitoring module, and an abnormal operation self-check cycle adjustment module:

[0009] The standby unit self-test and status monitoring module, first, executes a preset self-test and status monitoring cycle to perform regular tests on the standby unit to ensure that the standby unit is in normal working condition when in standby mode;

[0010] The real-time data acquisition module of the main control unit collects various operating parameters of the main control unit in real time while the standby unit is performing regular inspections, providing information basis for subsequent preprocessing and feature extraction;

[0011] The data preprocessing and quality improvement module preprocesses the acquired operating parameter information to improve data quality and reliability, and provide high-quality data support for subsequent feature extraction and analysis;

[0012] The key feature extraction and analysis module, after the data preprocessing is completed, extracts the key features reflecting the abnormal operation of the main control unit from the preprocessed data, analyzes the extracted key features under the detection window, and quantifies the current operation status of the main control unit;

[0013] The deep learning model prediction module inputs the analyzed key features into the pre-learned deep learning model, and uses the deep learning model to intelligently predict the operating status of the main control unit;

[0014] The operation status classification and decision-making module divides the operation status of the main control unit into two categories: normal operation and abnormal operation according to the prediction results of the deep learning model;

[0015] The routine maintenance and monitoring module, for normal operation, continues to use the preset periodic self-test and status detection cycle to maintain the routine maintenance frequency of the standby unit;

[0016] The abnormal operation self-check cycle adjustment module reduces the self-check and status detection cycle and increases the self-check frequency for abnormal operation status, ensuring that the backup unit can quickly and reliably take over the control task when an abnormality occurs in the main control unit.

[0017] Preferably, after data preprocessing is completed, key features reflecting abnormal operation of the main control unit are extracted from the preprocessed data, the extracted features include the degree of response of the main control unit to the input signal and the power fluctuation when the main control unit responds to the control command, and under the detection window, the degree of response of the main control unit to the input signal and the power fluctuation when the main control unit responds to the control command are analyzed to generate a response nonlinear factor and a response power fluctuation factor, respectively. The response nonlinear factor quantifies the nonlinear characteristics of the degree of response of the main control unit to the input signal, and the response power fluctuation factor quantifies the degree of power fluctuation generated by the main control unit when executing the control command.

[0018] Preferably, the specific steps of analyzing the response degree of the main control unit to the input signal in the detection window to generate the response nonlinear factor are as follows:

[0019] First, preprocess the input signal to reduce signal interference. The expression is as follows:

[0020]

[0021] , where u′(t) is the preprocessed input signal, u(t) is the original input signal, min(u(t)) is the minimum value of the original input signal, and max(u(t)) is the maximum value of the original input signal;

[0022] After the input signal is preprocessed, a nonlinear response analysis model is constructed by establishing a mapping relationship between the preprocessed input signal and the output signal of the control unit. The relationship between the input signal and the output signal u′(t) is as follows:

[0023] y(t)=αu′(t)+βu′(t) 2 +γu′(t) 3 +∈u′(t) n

[0024] , where y(t) is the output signal, α is the linear response coefficient, β is the quadratic nonlinear response coefficient, γ is the cubic nonlinear response coefficient, ∈ is the higher-order nonlinear response coefficient, and n is the nonlinear order;

[0025] The core purpose of the response nonlinear factor is to measure the nonlinear response degree of the input signal and quantify the abnormality of this response. The abnormality is quantified by calculating the nonlinear residual coefficient. The calculation expression is as follows:

[0026]

[0027] , where NLR is the nonlinear residual coefficient, is the predicted output signal of the nonlinear response analysis model, is the median of the output signal, T is the detection window length;

[0028] Finally, by comprehensively evaluating the nonlinear residual coefficient NLR and introducing the time weighting function to enhance the sensitivity of the abnormal state, the response nonlinear factor is generated. The generation formula is as follows:

[0029]

[0030] , where Response Non is the response nonlinear factor, NLR(t) is the nonlinear residual coefficient at time point t, and W(t) is the time weight function.

[0031] Preferably, the specific steps of analyzing the power fluctuation when the main control unit responds to the control command in the detection window to generate the response power fluctuation factor are as follows:

[0032] The collected control command response power signal is preprocessed, and the instantaneous power fluctuation rate of the signal is calculated after obtaining the power fluctuation data. The calculation expression is as follows:

[0033]

[0034] , where P filtered (t) is the power fluctuation signal after preprocessing, Δt is the time interval, ΔP(t) is the instantaneous power fluctuation rate, and t is the time point;

[0035] The nonlinear fluctuation response characteristics of the calculated signal are further quantified to further quantify the responsiveness of the main control unit to the control command. The nonlinear characteristics are obtained by Fourier transforming the instantaneous power fluctuation rate. The calculation expression is as follows:

[0036]

[0037] , where It is the nonlinear response characteristic generated by the instantaneous power fluctuation rate ΔP(t) after Fourier transformation, e is the natural base, j is the imaginary unit, and f is the frequency;

[0038] Finally, according to the obtained nonlinear fluctuation response characteristics, in order to quantify the abnormal degree of power fluctuation, a nonlinear exponential function is introduced to combine the frequency component in the signal with the time response to generate the response power fluctuation factor. The generation formula is as follows:

[0039]

[0040] , where Response Power is the response power fluctuation factor, exp(-ωf 2 ) is the exponential decay factor, and ω is the exponential decay constant, which controls the speed of the decay factor.

[0041] Preferably, the response nonlinear factor and the response power fluctuation factor generated after analyzing the extracted key features are input into a pre-learned deep learning model, and an operation fluctuation index is generated by the deep learning model, and the operation status of the main control unit is intelligently predicted by the operation fluctuation index.

[0042] Preferably, the operation fluctuation index generated when the operation state of the main control unit is intelligently predicted by the pre-learned deep learning model is compared and analyzed with the pre-set reference threshold of the operation fluctuation index, and the operation state of the main control unit is divided, and the division steps are as follows:

[0043] If the operation fluctuation index is greater than a preset operation fluctuation index reference threshold, the operation state of the main control unit is classified as abnormal operation;

[0044] If the operation fluctuation index is less than or equal to a preset operation fluctuation index reference threshold, the operation state of the main control unit is classified as normal operation.

[0045] Preferably, for abnormal operation status, the specific steps of reducing the self-check and status detection cycle, increasing the self-check frequency, and ensuring that the backup unit can quickly and reliably take over the control task when the main control unit is abnormal are as follows:

[0046] First, based on the operational volatility index Fluc Determine the current operating status of the main control unit. Fluc When the preset operation fluctuation index reference threshold is exceeded, it means that the main control unit has entered an abnormal operation state, and the backup unit must quickly perform self-check to ensure its normal operation. The calculation expression is as follows:

[0047]

[0048] , where Operational Fluc Operational Volatility Index ref is the reference threshold of the operating fluctuation index, T preset is the preset periodic self-check and status check cycle, T backup It is the dynamic self-test cycle of the standby unit;

[0049] After adjusting the self-test cycle according to the operation fluctuation index, according to the dynamic self-test cycle T of the standby unit backup To further adjust the self-test frequency of the standby unit, the self-test frequency adjustment formula is as follows:

[0050]

[0051] , where F backup is the self-test frequency of the standby unit, θ is the adjustment factor;

[0052] In order to ensure the reliability of the backup unit, the self-test frequency is increased and the depth of the self-test is dynamically adjusted according to the severity of the fluctuation. The formula for adjusting the self-test depth is as follows:

[0053]

[0054] , where D backup is the self-test depth adjustment factor of the standby unit, D preset It is the preset standard self-test depth coefficient; is the depth adjustment factor;

[0055] Finally, according to the frequency changes of the standby unit self-check and status monitoring, the standby unit response time is optimized to ensure that when the main control unit is abnormal, it can respond quickly and complete the switch. The optimization formula is as follows:

[0056]

[0057] , where T response is the response time of the optimized backup unit, T base is the backup unit basic response time.

[0058] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0059] The present invention ensures that the standby unit is always in a normal working state when on standby by regularly performing self-inspection and status monitoring of the standby unit, and can effectively prevent the standby unit from failing due to long-term non-detection. In addition, by real-time monitoring of the operating status of the main control unit, combined with data preprocessing, feature extraction and deep learning prediction models, it is possible to dynamically evaluate whether the main control unit is in an abnormal state. If an abnormality occurs in the main control unit, the self-inspection and status monitoring cycle of the standby unit is automatically identified and adjusted in advance to ensure that the standby unit can quickly and reliably take over the control task when the main control unit fails. In particular, when there is a slight fluctuation in the main control unit, the self-inspection frequency of the standby unit is increased, reducing the risk of failure of the standby unit not being discovered in time, greatly improving the success rate of redundant switching, and avoiding the failure of redundant protection due to failure of the standby unit, thereby reducing the probability of failure and production interruption.

[0060] The present invention uses a deep learning model to intelligently predict the operating status of the main control unit, and can adjust the maintenance strategy of the backup unit in real time according to the abnormal performance of the main control unit. When the main control unit is in normal operation, the preset regular self-test and status detection cycle continues to be used to maintain the regular maintenance frequency of the backup unit, ensuring that the backup unit does not cause excessive waste of resources when on standby. When an abnormality occurs in the main control unit, the backup unit will quickly reduce the self-test cycle and increase the self-test frequency of the backup unit to ensure that it is in good working condition when taking over the task. Through this intelligent and adaptive maintenance mechanism, the situation of excessive maintenance or neglect of maintenance of the backup unit is avoided, which not only improves the reliability of the backup unit, but also effectively saves maintenance costs, reduces unnecessary downtime and production losses, and ultimately improves the stability and economic benefits of the entire system. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced 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.

[0062] Figure 1 It is a module schematic diagram of the redundant control system of the coal feeder based on DCS and SCM of the present invention. DETAILED DESCRIPTION

[0063] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0064] The present invention provides Figure 1 The redundant control system of the coal feeder based on DCS and SCM shown in the figure includes a standby unit self-check and status monitoring module, a main control unit real-time data acquisition module, a data preprocessing and quality improvement module, a key feature extraction and analysis module, a deep learning model prediction module, an operation status classification and decision module, a routine maintenance and monitoring module, and an abnormal operation self-check cycle adjustment module:

[0065] The standby unit self-test and status monitoring module, first, executes a preset self-test and status monitoring cycle to perform regular tests on the standby unit to ensure that the standby unit is in normal working condition when in standby mode;

[0066] The main purpose of this step is to ensure that the backup unit is always in normal working condition when on standby. Through the preset periodic inspection, hardware failures (such as processor failure, power supply problems, memory corruption) or software failures (such as control program errors) in the backup unit can be discovered in time and repaired or replaced before the failure affects the system. This mechanism is the basis of the entire redundant control system, ensuring that the backup unit has the ability to take over tasks when needed, thereby ensuring the high availability and stability of the system.

[0067] The real-time data acquisition module of the main control unit collects various operating parameters of the main control unit in real time while the standby unit is performing regular inspections, providing information basis for subsequent preprocessing and feature extraction;

[0068] These parameters may include key indicators such as temperature, pressure, flow, speed, and current, reflecting the operating status and environmental conditions of the main control unit. Through real-time data acquisition, the system can continuously monitor the operating status of the main control unit and promptly detect any abnormal fluctuations or potential failures. These data provide the necessary information basis for subsequent preprocessing and feature extraction, and are the prerequisite for realizing intelligent prediction and dynamic adjustment.

[0069] The data preprocessing and quality improvement module preprocesses the acquired operating parameter information to improve data quality and reliability, and provide high-quality data support for subsequent feature extraction and analysis;

[0070] The collected real-time operating parameters often contain noise, outliers or incomplete data, so preprocessing is required to improve data quality and reliability. The preprocessing steps include data cleaning, denoising, normalization and standardization to ensure data consistency and accuracy. Through preprocessing, the system can eliminate interference factors and extract meaningful signals, thereby providing high-quality data support for subsequent feature extraction and analysis. This step ensures the accuracy of data analysis and the effectiveness of deep learning models.

[0071] The key feature extraction and analysis module, after the data preprocessing is completed, extracts the key features reflecting the abnormal operation of the main control unit from the preprocessed data, analyzes the extracted key features under the detection window, and quantifies the current operation status of the main control unit;

[0072] After data preprocessing is completed, key features reflecting abnormal operation of the main control unit are extracted from the preprocessed data. The extracted features include the response degree of the main control unit to the input signal and the power fluctuation when the main control unit responds to the control command. Under the detection window, the response degree of the main control unit to the input signal and the power fluctuation when the main control unit responds to the control command are analyzed to generate a response nonlinear factor and a response power fluctuation factor, respectively. The response nonlinear factor quantifies the nonlinear characteristics of the response degree of the main control unit to the input signal, and the response power fluctuation factor quantifies the power fluctuation degree generated by the main control unit when executing the control command.

[0073] A nonlinear response of the main control unit to the input signal usually indicates that the system is in an abnormal operating state. Under normal circumstances, the main control unit should respond linearly to the input signal, that is, there is a stable and predictable relationship between the input and the output. If the main control unit responds to the input signal nonlinearly, it means that the control algorithm or hardware of the system has unstable behavior and cannot perform the correct signal processing and feedback as expected. Nonlinear response can be caused by a variety of factors, including errors in the control system algorithm, sensor failure, hardware performance degradation (such as processor failure or memory problems), or excessive system load. For example, under overload conditions, the control unit may not be able to process the input signal in time, causing the output signal to be distorted or fluctuate violently, thus exhibiting nonlinear behavior. At this point, the control ability and response accuracy of the system are greatly reduced, and it cannot provide a stable output, which may cause equipment damage, process deviations, and even production stagnation. Therefore, nonlinear response is an important indicator of abnormal operation, which can help detect and diagnose potential faults or system instability, and take preventive measures in time to avoid further losses.

[0074] The specific steps of analyzing the response degree of the main control unit to the input signal under the detection window to generate the response nonlinear factor are as follows:

[0075] First, the input signal is preprocessed. Since the input signal may be subject to noise or external interference in actual operation, it is necessary to preprocess the signal to reduce signal interference. The expression is as follows:

[0076]

[0077] , where u′(t) is the preprocessed input signal, u(t) is the original input signal, min(u(t)) is the minimum value of the original input signal, and max(u(t)) is the maximum value of the original input signal;

[0078] The preprocessed input signal u′(t) is easier to compare with the output signal and helps to remove unnecessary dimensional differences, providing a standardized input data for subsequent analysis.

[0079] After the input signal is preprocessed, a nonlinear response analysis model is constructed by establishing a mapping relationship between the preprocessed input signal and the output signal of the control unit. In order to quantify the degree of nonlinearity between input and output, a nonlinear response model based on Kerr's law is used, which can effectively capture the nonlinear dynamic characteristics. The relationship between the input signal and the output signal u′(t) is as follows:

[0080] y(t)=αu′(t)+βu′(t) 2 +γu′(t) 3 +∈u′(t) n

[0081] , where y(t) is the output signal, α is the linear response coefficient, which represents the linear relationship coefficient between the preprocessed input signal u′(t) and the output signal y(t), β is the quadratic nonlinear response coefficient, which represents the influence of the square term of the preprocessed input signal u′(t) on the output signal y(t), γ is the cubic nonlinear response coefficient, which represents the influence of the cubic term of the preprocessed input signal u′(t) on the output signal y(t), and ∈ is the high-order nonlinear response coefficient, which represents the high-order term of the input signal u′(t) n The impact on the output signal y(t), n is the nonlinear order, which indicates the nonlinear degree of the main control unit response;

[0082] The core of the model lies in the construction of polynomials, in which the high-order power terms of the preprocessed input signal u′(t) (such as u′(t) 3 and u′(t) n ) can capture the nonlinear behavior in the system. By fitting the relationship between the input signal and the output signal using the least squares method, the parameters in the above model can be obtained, and then the nonlinear response characteristics of the main control unit can be quantified.

[0083] The core purpose of the response nonlinear factor is to measure the nonlinear response degree of the input signal and quantify the abnormality of this response. The abnormality is quantified by calculating the nonlinear residual coefficient. The nonlinear residual coefficient is based on the difference between the model predicted output and the actual output, reflecting the nonlinear response degree and abnormal fluctuation of the system to the input signal. The calculation expression is as follows:

[0084]

[0085] , where NLR is the nonlinear residual coefficient, is the predicted output signal of the nonlinear response analysis model, is the median of the output signal, T is the detection window length;

[0086] The median value of the output signal It refers to the value in the middle of the detection window after the values ​​are arranged in order. If there are an even number of data points in the detection window, the average of the two middle data points is taken. The median is a statistic that measures the trend of the data set. It is more resistant to the influence of extreme outliers than the average value, especially when the data distribution is asymmetric or there is noise, it can more accurately reflect the center position of the signal. By calculating the median of the output signal We can exclude extreme high or low values ​​to ensure that the analysis process is not disturbed by occasional abnormal fluctuations, and further help quantify the normal operation state of the system and the degree of deviation from the nonlinear response. In practical applications, the median of the output signal is often used to measure the steady-state behavior of the system, providing a more robust benchmark for subsequent nonlinear analysis.

[0087] Finally, by comprehensively evaluating the nonlinear residual coefficient NLR and introducing the time weighting function to enhance the sensitivity of the abnormal state, the response nonlinear factor is generated. The generation formula is as follows:

[0088]

[0089] , where Response Non is the response nonlinear factor, NLR(t) is the nonlinear residual coefficient at time point t, and W(t) is the time weight function, which weights the nonlinear residual coefficient NLR(t) at each time point t.

[0090] The larger the response nonlinear factor value generated after analyzing the response of the main control unit to the input signal under the detection window, the more abnormal the operation of the main control unit may be. Under normal circumstances, the main control unit should show a linear response to the input signal, that is, a stable proportional relationship between the input and the output. If the response nonlinear factor value is large during analysis within the monitoring window, it means that the main control unit has obvious nonlinear behavior in response to the input signal, which may be due to a fault or unstable operation of the control system, such as an error in the control algorithm, a hardware failure, or an excessive system load. If the value of the response nonlinear factor is small or close to zero, it means that the main control unit remains linear in response to the input signal, indicating that the main control unit is operating normally.

[0091] When the main control unit responds to the control command, the power fluctuates significantly, which usually indicates that the main control unit is in an abnormal operating state. Under normal circumstances, after receiving the control command, the main control unit should stably adjust the power output according to the system requirements to ensure the smooth operation of the equipment or load. However, when the main control unit fails or the load is abnormal, the response power fluctuation may become violent and irregular, which usually indicates that there is a problem with the control process of the system. Such significant power fluctuations may be caused by a variety of factors, such as errors in the control algorithm, hardware failures (such as processor overload, memory damage, unstable power supply) or external interference (such as abnormal fluctuations or noise in the input signal). In some cases, the main control unit may not be able to adjust the output power in time to adapt to the load demand, resulting in unstable power, or even overload or insufficient power. At this time, the main control unit fails to effectively maintain the control target, which seriously affects the stability and safety of the system.

[0092] The specific steps of analyzing the power fluctuation when the main control unit responds to the control command in the detection window to generate the response power fluctuation factor are as follows:

[0093] The collected control command response power signal is preprocessed, and the instantaneous power fluctuation rate of the signal is calculated after obtaining the power fluctuation data. The instantaneous power fluctuation rate reflects the rapid change rate of power output and helps capture the small fluctuations when the control unit responds. The calculation expression is as follows:

[0094]

[0095] , where P filtered (t) is the power fluctuation signal after preprocessing, Δt is the time interval, ΔP(t) is the instantaneous power fluctuation rate, which represents the rate of change of the power output signal of the main control unit within the time interval Δt, specifically the difference between two adjacent sampling values ​​of the power signal, and t is the time point;

[0096] A large ΔP(t) indicates a sharp change in power during the control response, which usually indicates abnormal system behavior, such as oscillation or sudden fluctuations in the control signal.

[0097] The nonlinear fluctuation response characteristics of the signal are calculated to further quantify the responsiveness of the main control unit to the control command. The nonlinear characteristics are obtained by Fourier transforming the instantaneous power fluctuation rate. This process helps to reveal the potential nonlinear behavior and periodic fluctuations in the signal. The calculation expression is as follows:

[0098]

[0099] , where It is the nonlinear response characteristic generated by the instantaneous power fluctuation rate ΔP(t) after Fourier transformation, e is the natural base, j is the imaginary unit, and f is the frequency;

[0100] The spectrum features extracted by Fourier transform can reveal the high-frequency fluctuation components in the signal and show whether the system has frequent fluctuations or over-adjustments in a short period of time.

[0101] Finally, according to the obtained nonlinear fluctuation response characteristics, in order to quantify the abnormal degree of power fluctuation, a nonlinear exponential function is introduced to combine the frequency component in the signal with the time response to generate the response power fluctuation factor. The generation formula is as follows:

[0102]

[0103] , where Response Power is the response power fluctuation factor, exp(-ωf 2 ) is the exponential decay factor applied to each frequency component, and ω is the exponential decay constant that controls the speed of the decay factor.

[0104] The exponential decay factor is a mathematical function of the form exp(-ωf 2 ), where ω is a constant and f is the frequency. This factor describes how some signal or feature gradually becomes less important in the calculation as the frequency increases. Specifically, the exponential decay factor acts to weight the frequency components in the signal, where lower frequency components contribute more to the result, while the effect of higher frequency components decays rapidly. Its role is to reduce the impact of high-frequency components and emphasize the contribution of low-frequency fluctuations to the behavior of the system, which is very important in many practical applications, especially when high-frequency fluctuations may come from noise or irrelevant short-term fluctuations. In this way, the exponential decay factor helps focus on the more representative and stable low-frequency fluctuations, thereby enhancing the system's responsiveness to long-term, stable features, while reducing the interference of high-frequency noise on the analysis results.

[0105] The larger the performance value of the response power fluctuation factor generated after analyzing the power fluctuation when the main control unit responds to the control command under the detection window, the higher the performance value, which usually indicates that the main control unit is in an abnormal operating state. The larger the value of the power fluctuation factor, the more violent and unstable the power fluctuation of the main control unit when responding to the control command, which may be caused by a variety of reasons, such as excessive load on the main control unit, control algorithm problems, power supply failure or hardware damage. This abnormal fluctuation usually reflects that the main control unit has failed to stably adjust the power output as expected, resulting in abnormal or unstable operation of the main control unit. On the contrary, if the performance value of the response power fluctuation factor is small, it means that the power fluctuation is relatively stable, and the main control unit can stably and reliably respond to the control command and maintain normal power output, indicating that the main control unit is in normal operation.

[0106] The deep learning model prediction module inputs the analyzed key features into the pre-learned deep learning model, and uses the deep learning model to intelligently predict the operating status of the main control unit;

[0107] The response nonlinear factor and response power fluctuation factor generated after analyzing the extracted key features are input into the pre-learned deep learning model, and the operation fluctuation index is generated by the deep learning model. The operation status of the main control unit is intelligently predicted by the operation fluctuation index.

[0108] A pre-learned deep learning model is a model that is trained and optimized by deep learning algorithms (such as neural networks, convolutional neural networks, long short-term memory networks, etc.) based on a large amount of historical data and known control unit operation modes before the system is put into actual operation. These historical data usually include input and output data under normal operation, various system parameters, and feature markers during abnormal operation (such as key performance changes before and after a fault occurs), and these data are used to identify the laws of system behavior. The training process of a deep learning model involves continuously optimizing the weights and parameters of the model through algorithms, so that it can effectively learn and extract implicit features and patterns from the input data, so as to predict the future state of the system. During the training process, the model uses the "supervised learning" method to learn using labeled data (such as labels for normal and abnormal states), and gradually adjusts the structure of the model to accurately identify and predict system state changes.

[0109] Through the learning of the deep learning model, the model can use the learned features and patterns to make inferences and predictions when faced with new and unseen data. In practical applications, the pre-learned deep learning model can input key features extracted from the main control unit, such as the response nonlinear factor and the response power fluctuation factor, to make intelligent predictions on the operating status of the system. Specifically, the deep learning model will judge whether the main control unit is in normal operation or whether there is a potential abnormal risk based on these features. Ultimately, the operation fluctuation index generated by the model provides a quantitative indicator for the system, through which the operating stability and health of the main control unit can be dynamically evaluated. If the operation fluctuation index shows abnormal fluctuations or growth, the system can respond in time, such as adjusting the self-test cycle, increasing fault monitoring and other measures, to avoid more serious failures or production interruptions.

[0110] This process reflects the application advantages of deep learning in industrial control systems, especially in complex nonlinear systems and big data environments, where deep learning models can automatically learn and extract key features from large amounts of real-time data without human intervention. Through this intelligent prediction, the system can identify potential problems in advance and issue warnings or handle them before they actually affect the system, greatly improving the reliability and stability of the system.

[0111] The deep learning model is not limited here and can respond to nonlinear factors over time. Non and response power fluctuation factor Response Power Conduct comprehensive analysis to generate operational volatility index Fluc In order to realize the technical solution of the present invention, the present invention provides a specific implementation method;

[0112] Operational Fluc The generation formula is as follows:

[0113]

[0114] , where f a 、f b They are the response nonlinear factors Response Non and response power fluctuation factor Response Power The preset proportionality factor, and f a 、f b Both are greater than 0.

[0115] It can be seen from the operation fluctuation index that the larger the response nonlinear factor expression value generated after analyzing the response degree of the main control unit to the input signal under the detection window, the larger the response power fluctuation factor expression value generated after analyzing the power fluctuation when the main control unit responds to the control command under the detection window, and the larger the operation fluctuation index expression value generated when the operating state of the main control unit is intelligently predicted by the pre-learned deep learning model, the greater the probability of abnormality occurring during the operation of the main control unit, and vice versa, the smaller the probability of abnormality occurring during the operation of the main control unit.

[0116] The operation status classification and decision-making module divides the operation status of the main control unit into two categories: normal operation and abnormal operation according to the prediction results of the deep learning model;

[0117] The operation fluctuation index generated by intelligently predicting the operation status of the main control unit through the pre-learned deep learning model is compared and analyzed with the pre-set reference threshold of the operation fluctuation index to divide the operation status of the main control unit. The division steps are as follows:

[0118] If the operation fluctuation index is greater than a preset operation fluctuation index reference threshold, the operation state of the main control unit is classified as abnormal operation;

[0119] If the operation fluctuation index is less than or equal to a preset operation fluctuation index reference threshold, the operation state of the main control unit is classified as normal operation.

[0120] Normal operation means that the main control unit can continuously and stably perform control tasks within the predetermined working range, meet performance requirements, and maintain various system parameters (such as temperature, pressure, power, etc.) within the normal working range. Abnormal operation means that the operating state of the main control unit deviates from normal expectations, which is manifested in unstable operation of the main control unit, failure to execute control instructions as expected, or unpredictable power fluctuations, response delays, frequent errors or failures, etc.

[0121] The routine maintenance and monitoring module, for normal operation, continues to use the preset periodic self-test and status detection cycle to maintain the routine maintenance frequency of the standby unit;

[0122] For normal operation, continue to use the preset regular self-test and status detection cycle and maintain the regular maintenance frequency of the backup unit to ensure that the backup unit is always in the best working condition and is ready to take over the control task at any time. Through regular self-test and status detection, potential hardware failures, software anomalies or performance degradation can be discovered and eliminated in a timely manner, ensuring that the backup unit can seamlessly take over when the main control unit fails, avoiding the failure of redundant switching due to failure or unstable state of the backup unit itself. Regular maintenance helps to extend the service life of the backup unit, reduce the risk of failure in long-term non-use, ensure the high reliability and stability of the system, and especially in the face of emergencies, to ensure the continuity and safety of production.

[0123] Abnormal operation self-check cycle adjustment module, for abnormal operation status, reduces the self-check and status detection cycle, increases the self-check frequency, and ensures that the backup unit can quickly and reliably take over the control task when the main control unit is abnormal;

[0124] For abnormal operation status, reduce the self-test and status detection cycle, increase the self-test frequency, and ensure that the backup unit can quickly and reliably take over the control task when the main control unit fails. The specific steps are as follows:

[0125] First, based on the operational volatility index Fluc Determine the current operating status of the main control unit. Fluc When the preset operation fluctuation index reference threshold is exceeded, it means that the main control unit has entered an abnormal operation state, and the backup unit must quickly perform self-test to ensure its normal operation. At this time, the self-test cycle of the backup unit needs to be dynamically adjusted according to the actual fluctuation situation. The calculation expression is as follows:

[0126]

[0127] , where Operational Fluc Operational Volatility Index ref is the reference threshold of the operating fluctuation index, T preset is the preset periodic self-check and status check cycle, T backup It is the dynamic self-test cycle of the standby unit;

[0128] If Operational Fluc >Operational ref , then T backup will be less than T preset , thereby shortening the self-test interval of the backup unit, ensuring that the backup unit can be checked in a shorter time and be ready to take over the control task at any time.

[0129] After adjusting the self-test cycle according to the operation fluctuation index, according to the dynamic self-test cycle T of the standby unit backup To further adjust the self-test frequency of the standby unit, the self-test frequency is inversely proportional to the cycle. The shorter the cycle, the higher the self-test frequency should be. The self-test frequency adjustment formula is as follows:

[0130]

[0131] , where F backup is the self-test frequency of the standby unit, θ is the adjustment factor used to control the sensitivity of frequency changes, usually a constant less than 1 to prevent over-adjustment;

[0132] Operational Volatility Index Fluc Increase, the dynamic self-test period T of the standby unit backup Reduce, so that the standby unit self-test frequency F backup The increased frequency means that the backup unit will self-check more frequently, ensuring a quick response to abnormal conditions in the main control unit.

[0133] In order to ensure the reliability of the backup unit, the self-check frequency is increased and the depth of the self-check is dynamically adjusted according to the severity of the fluctuation. The depth adjustment is to analyze the abnormal fluctuation pattern of the main control unit, increase the inspection items of the backup unit for hardware and software, and reduce the inspection part that relies on the standard self-check procedure. The formula for adjusting the self-check depth is as follows:

[0134]

[0135] , where D backup is the self-test depth adjustment factor of the standby unit, which indicates the degree of detailed self-test performed by the standby unit under abnormal conditions. preset It is the preset standard self-test depth coefficient; is the depth adjustment factor, which determines the degree to which changes in the volatility index affect the depth of the self-test;

[0136] When running the volatility index Operational Fluc When the threshold is exceeded, the self-test depth of the backup unit is adjusted by the coefficient D backup As the power increases, the backup unit will perform more complex and in-depth self-test procedures, which helps to discover hidden hardware or software faults and ensures that the backup unit can take over the control task promptly and reliably when the main control unit is abnormal.

[0137] Finally, according to the frequency changes of the standby unit self-check and status monitoring, the standby unit response time is optimized to ensure that when the main control unit is abnormal, it can respond quickly and complete the switch. The response time of the standby unit is related to the self-check cycle, depth and frequency, and needs to be dynamically adjusted to adapt to operation fluctuations. The optimization formula is as follows:

[0138]

[0139] , where T response is the response time of the optimized backup unit, T base It is the basic response time of the backup unit, indicating the response time of the backup unit in the absence of abnormalities.

[0140] The above steps show that as the frequency and depth of self-checking increase, the response time of the backup unit will be optimized accordingly. By adjusting these parameters, redundancy switching can be completed quickly when the main control unit is abnormal, ensuring the continuity of production and the safety of the equipment.

[0141] In abnormal operation, measures to reduce the self-test and status detection cycle and increase the self-test frequency are aimed at ensuring that the backup unit can quickly and reliably take over the control task when the main control unit fails or fluctuates. The core function of this measure is to dynamically adjust the self-test strategy of the backup unit so that it can respond quickly to changes in the operating status of the main control unit, thereby ensuring that the backup unit is in good working condition at any time when the system is abnormal.

[0142] Specifically, when the operation fluctuation index of the main control unit exceeds the preset operation fluctuation index reference threshold, the main control unit enters an abnormal operation state, and the backup unit needs to be quickly prepared to take over the control task. In this case, the traditional regular self-check cycle may not be able to detect potential failures of the backup unit in a timely manner, thereby affecting the reliability of redundant protection. By shortening the self-check cycle and increasing the self-check frequency, a comprehensive inspection can be performed in a shorter time to ensure that the backup unit is in the best working condition when the system is abnormal. The quantitative adjustment of this process is dynamically evaluated through the operation fluctuation index to ensure that the self-check frequency and depth of the backup unit are adjusted in real time according to the degree of abnormality of the main control unit, avoiding the failure of redundant switching due to failure of the backup unit itself.

[0143] Through this measure, when the main control unit fluctuates or fails, the backup unit can quickly perform self-checks and take over tasks, minimizing production line downtime and ensuring the continuity and safety of the production process. In addition, this can also avoid irreversible system failures caused by insufficient inspection of the backup unit, thereby improving the reliability and stability of the entire system.

[0144] The present invention ensures that the standby unit is always in a normal working state when on standby by regularly performing self-inspection and status monitoring of the standby unit, and can effectively prevent the standby unit from failing due to long-term non-detection. In addition, by real-time monitoring of the operating status of the main control unit, combined with data preprocessing, feature extraction and deep learning prediction models, it is possible to dynamically evaluate whether the main control unit is in an abnormal state. If an abnormality occurs in the main control unit, the self-inspection and status monitoring cycle of the standby unit is automatically identified and adjusted in advance to ensure that the standby unit can quickly and reliably take over the control task when the main control unit fails. In particular, when there is a slight fluctuation in the main control unit, the self-inspection frequency of the standby unit is increased, reducing the risk of failure of the standby unit not being discovered in time, greatly improving the success rate of redundant switching, and avoiding the failure of redundant protection due to failure of the standby unit, thereby reducing the probability of failure and production interruption.

[0145] The present invention uses a deep learning model to intelligently predict the operating status of the main control unit, and can adjust the maintenance strategy of the backup unit in real time according to the abnormal performance of the main control unit. When the main control unit is in normal operation, the preset regular self-test and status detection cycle continues to be used to maintain the regular maintenance frequency of the backup unit, ensuring that the backup unit does not cause excessive waste of resources when on standby. When an abnormality occurs in the main control unit, the backup unit will quickly reduce the self-test cycle and increase the self-test frequency of the backup unit to ensure that it is in good working condition when taking over the task. Through this intelligent and adaptive maintenance mechanism, the situation of excessive maintenance or neglect of maintenance of the backup unit is avoided, which not only improves the reliability of the backup unit, but also effectively saves maintenance costs, reduces unnecessary downtime and production losses, and ultimately improves the stability and economic benefits of the entire system.

[0146] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0147] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0148] It should be noted that, in this article, if there are first and second, etc., relational terms are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0149] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0150] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0151] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0152] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0153] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0154] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0155] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. The redundant control system of coal feeder based on DCS and SCM is characterized by: It includes standby unit self-check and status monitoring module, main control unit real-time data acquisition module, data preprocessing and quality improvement module, key feature extraction and analysis module, deep learning model prediction module, operation status classification and decision module, routine maintenance and monitoring module and abnormal operation self-check cycle adjustment module: The standby unit self-test and status monitoring module, first, executes a preset self-test and status monitoring cycle to perform regular tests on the standby unit to ensure that the standby unit is in normal working condition when in standby mode; The real-time data acquisition module of the main control unit collects various operating parameters of the main control unit in real time while the standby unit is performing regular inspections, providing information basis for subsequent preprocessing and feature extraction; The data preprocessing and quality improvement module preprocesses the acquired operating parameter information to improve data quality and reliability, and provide high-quality data support for subsequent feature extraction and analysis; The key feature extraction and analysis module, after the data preprocessing is completed, extracts the key features reflecting the abnormal operation of the main control unit from the preprocessed data, analyzes the extracted key features under the detection window, and quantifies the current operation status of the main control unit; The deep learning model prediction module inputs the analyzed key features into the pre-learned deep learning model, and uses the deep learning model to intelligently predict the operating status of the main control unit; The operation status classification and decision-making module divides the operation status of the main control unit into two categories: normal operation and abnormal operation according to the prediction results of the deep learning model; The routine maintenance and monitoring module, for normal operation, continues to use the preset periodic self-test and status detection cycle to maintain the routine maintenance frequency of the standby unit; The abnormal operation self-check cycle adjustment module reduces the self-check and status detection cycle and increases the self-check frequency for abnormal operation status, ensuring that the backup unit can quickly and reliably take over the control task when an abnormality occurs in the main control unit.

2. The redundant control system for coal feeder based on DCS and SCM according to claim 1 is characterized in that: After data preprocessing is completed, key features reflecting abnormal operation of the main control unit are extracted from the preprocessed data. The extracted features include the response degree of the main control unit to the input signal and the power fluctuation when the main control unit responds to the control command. Under the detection window, the response degree of the main control unit to the input signal and the power fluctuation when the main control unit responds to the control command are analyzed to generate a response nonlinear factor and a response power fluctuation factor, respectively. The response nonlinear factor quantifies the nonlinear characteristics of the response degree of the main control unit to the input signal, and the response power fluctuation factor quantifies the power fluctuation degree generated by the main control unit when executing the control command.

3. The redundant control system for coal feeder based on DCS and SCM according to claim 2 is characterized in that: The specific steps of analyzing the response degree of the main control unit to the input signal under the detection window to generate the response nonlinear factor are as follows: First, preprocess the input signal to reduce signal interference. The expression is as follows: ， Where u′(t) is the preprocessed input signal, u(t) is the original input signal, min(u(t)) is the minimum value of the original input signal, and max(u(t)) is the maximum value of the original input signal; After the input signal is preprocessed, a nonlinear response analysis model is constructed by establishing a mapping relationship between the preprocessed input signal and the output signal of the control unit. The relationship between the input signal and the output signal u′(t) is as follows: y(t)=αu′(t)+βu′(t) 2 +γu′(t) 3 +∈u′(t) n , Where y(t) is the output signal, α is the linear response coefficient, β is the quadratic nonlinear response coefficient, γ is the cubic nonlinear response coefficient, ∈ is the higher-order nonlinear response coefficient, and n is the nonlinear order; The core purpose of the response nonlinear factor is to measure the nonlinear response degree of the input signal and quantify the abnormality of this response. The abnormality is quantified by calculating the nonlinear residual coefficient. The calculation expression is as follows: ， Where NLR is the nonlinear residual coefficient, is the predicted output signal of the nonlinear response analysis model, is the median of the output signal, T is the detection window length; Finally, by comprehensively evaluating the nonlinear residual coefficient NLR and introducing the time weighting function to enhance the sensitivity of the abnormal state, the response nonlinear factor is generated. The generation formula is as follows: ， In the formula, Response Non is the response nonlinear factor, NLR(t) is the nonlinear residual coefficient at time point t, and W(t) is the time weight function.

4. The redundant control system for coal feeder based on DCS and SCM according to claim 2 is characterized in that: The specific steps of analyzing the power fluctuation when the main control unit responds to the control command in the detection window to generate the response power fluctuation factor are as follows: The collected control command response power signal is preprocessed, and the instantaneous power fluctuation rate of the signal is calculated after obtaining the power fluctuation data. The calculation expression is as follows: ， Where P filtered (t) is the power fluctuation signal after preprocessing, Δt is the time interval, ΔP(t) is the instantaneous power fluctuation rate, and t is the time point; The nonlinear fluctuation response characteristics of the calculated signal are further quantified to further quantify the responsiveness of the main control unit to the control command. The nonlinear characteristics are obtained by Fourier transforming the instantaneous power fluctuation rate. The calculation expression is as follows: ， In the formula, It is the nonlinear response characteristic generated by the instantaneous power fluctuation rate ΔP(t) after Fourier transformation, e is the natural base, j is the imaginary unit, and f is the frequency; Finally, according to the obtained nonlinear fluctuation response characteristics, in order to quantify the abnormal degree of power fluctuation, a nonlinear exponential function is introduced to combine the frequency component in the signal with the time response to generate the response power fluctuation factor. The generation formula is as follows: ， In the formula, Response Power is the response power fluctuation factor, exp(-ωf 2 ) is the exponential decay factor, and ω is the exponential decay constant, which controls the speed of the decay factor.

5. The redundant control system for coal feeder based on DCS and SCM according to claim 2 is characterized in that: The response nonlinear factor and response power fluctuation factor generated after analyzing the extracted key features are input into the pre-learned deep learning model, and the operation fluctuation index is generated by the deep learning model. The operation status of the main control unit is intelligently predicted by the operation fluctuation index.

6. The redundant control system for coal feeder based on DCS and SCM according to claim 5 is characterized in that: The operation fluctuation index generated by the pre-learned deep learning model when intelligently predicting the operation status of the main control unit is compared and analyzed with the pre-set reference threshold of the operation fluctuation index, and the operation status of the main control unit is divided. The division steps are as follows: If the operation fluctuation index is greater than a preset operation fluctuation index reference threshold, the operation state of the main control unit is classified as abnormal operation; If the operation fluctuation index is less than or equal to a preset operation fluctuation index reference threshold, the operation state of the main control unit is classified as normal operation.

7. The redundant control system for coal feeder based on DCS and SCM according to claim 6, characterized in that: For abnormal operation status, reduce the self-test and status detection cycle, increase the self-test frequency, and ensure that the backup unit can quickly and reliably take over the control task when the main control unit fails. The specific steps are as follows: First, based on the operational volatility index Fluc Determine the current operating status of the main control unit. Fluc When the preset operation fluctuation index reference threshold is exceeded, it means that the main control unit has entered an abnormal operation state, and the backup unit must quickly perform self-checks to ensure its normal operation. The calculation expression is as follows: ， In the formula, Operational Fluc Operational Volatility Index ref is the reference threshold of the operating fluctuation index, T preset is the preset periodic self-check and status check cycle, T backup It is the dynamic self-test cycle of the standby unit; After adjusting the self-test cycle according to the operation fluctuation index, according to the dynamic self-test cycle T of the standby unit backup To further adjust the self-test frequency of the standby unit, the self-test frequency adjustment formula is as follows: ， In the formula, F backup is the self-test frequency of the standby unit, θ is the adjustment factor; In order to ensure the reliability of the backup unit, the self-test frequency is increased and the depth of the self-test is dynamically adjusted according to the severity of the fluctuation. The formula for adjusting the self-test depth is as follows: ， Where D backup is the self-test depth adjustment factor of the standby unit, D preset It is the preset standard self-test depth coefficient; is the depth adjustment factor; Finally, according to the frequency changes of the standby unit self-check and status monitoring, the standby unit response time is optimized to ensure that when the main control unit is abnormal, it can respond quickly and complete the switch. The optimization formula is as follows: ， Where, T response is the response time of the optimized backup unit, T base is the backup unit basic response time.

Images