Method and system for suppressing back corona by regulating the timing of the rapping

Abstract

This application discloses a method and system for peak-shaving control of rapping timing to suppress back corona, relating to the field of process control technology. The method includes: equipping each dust-collecting electric field control zone of an electrostatic precipitator with a rapping actuator and an electrical parameter monitoring unit; real-time acquisition of secondary voltage and secondary current timing data for each rapping control zone; calculation of back corona severity indices; construction of a historical back corona intensity topology sequence for the dust-collecting electric field; prediction of the predicted back corona intensity sequence for each rapping control zone within a preset rapping control cycle; construction of a predicted back corona intensity topology sequence for the dust-collecting electric field; then performing rapping sequence optimization to solve for the optimal rapping timing scheme; and, according to the optimal rapping timing scheme, sequentially triggering the rapping actuators of each corresponding zone in chronological order. This solves the technical problem in existing technologies where unreasonable rapping timing leads to large secondary dust emissions and poor back corona suppression.

Description

Technical Field

[0001] This application relates to the field of process control technology, specifically to a method and system for controlling the timing of rapping to suppress back corona. Background Technology

[0002] With increasingly stringent environmental emission standards, the control requirements for pollutant emissions from coal-fired power plants are becoming more and more stringent. As the mainstream dust removal equipment, the dust removal performance of electrostatic precipitators directly determines whether the outlet dust emission concentration can meet the standards. Back corona is an abnormal phenomenon that is very likely to occur during the operation of electrostatic precipitators. When the dust accumulation thickness on the surface of the collecting or discharging electrodes is too large or the dust resistivity is too high, it will lead to difficulty in releasing the charge from the dust layer, forming a local reverse electric field. This can break down the dust layer and trigger back corona, which will not only significantly reduce dust collection efficiency but also increase energy consumption. In severe cases, it can cause the electrostatic precipitator to malfunction. Vibration cleaning is the main technical means to alleviate the back corona phenomenon. By vibrating, the dust accumulated on the plates and wires is removed, maintaining the normal operation of the electric field.

[0003] However, most existing rapping control strategies adopt fixed-sequence unified rapping or simultaneous rapping of all zones. This control method has obvious defects: on the one hand, simultaneous rapping of multiple areas will generate a large amount of secondary dust in a short period of time, significantly increasing the dust concentration at the outlet and thus affecting the dust removal effect; on the other hand, the rapping sequence is not arranged according to the actual back corona development of different zones, which can easily lead to the problem of lag when the back corona has developed to a serious level, or the problem of excessive dust removal when the back corona is mild. This not only fails to effectively suppress back corona, but also increases the wear and tear of the rapping mechanism and shortens the service life of the equipment. Summary of the Invention

[0004] This application provides a method and system for controlling the timing of rapping to suppress back corona, which solves the technical problem in the prior art where unreasonable rapping timing leads to large secondary dust emissions and poor back corona suppression effect.

[0005] The technical solution to the above-mentioned technical problems in this application is as follows: In a first aspect, this application provides a method for controlling the timing of rapping to suppress back corona discharge, the method comprising: The electrostatic precipitator's collecting electric field is spatially divided into several independent rapping control zones, and each zone is independently equipped with a rapping actuator and an electrical parameter monitoring unit. The electrical parameter monitoring unit collects the secondary voltage and secondary current time sequence data of each rapping control zone in real time during operation, calculates the back corona severity index, determines the historical back corona intensity sequence of each rapping control zone in the historical monitoring period, and constructs the historical back corona intensity topology sequence of the dust collection electric field. Based on the historical back corona intensity topology sequence, the predicted back corona intensity sequence of each of the rapping control zones within the preset rapping control cycle is obtained, and the predicted back corona intensity topology sequence of the dust collection electric field is constructed. Using the historical back corona intensity topology sequence, the predicted back corona intensity topology sequence, and the preset back corona intensity upper limit as inputs, with the rapping safety limit as a constraint, and with the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, the rapping sequence optimization is performed to solve for the optimal rapping timing scheme. According to the optimal vibration timing scheme, the vibration execution devices of each corresponding zone are triggered sequentially in time order.

[0006] Secondly, this application provides a rapping timing misalignment system for suppressing back corona, comprising: The partition configuration module is used to spatially divide the dust collection electric field of the electrostatic precipitator into several independent rapping control partitions, and to equip each partition with an independent rapping actuator and an electrical parameter monitoring unit. The back corona intensity assessment module is used to collect the secondary voltage time sequence data and secondary current time sequence data of each of the rapping control zones in real time during operation through the electrical parameter monitoring unit, calculate the back corona severity index, determine the historical back corona intensity sequence of each of the rapping control zones in the historical monitoring period, and construct the historical back corona intensity topology sequence of the dust collection electric field. The back corona intensity prediction module is used to predict and obtain the predicted back corona intensity sequence of each of the rapping control zones within a preset rapping control cycle based on the historical back corona intensity topology sequence, and to construct the predicted back corona intensity topology sequence of the dust collection electric field. The rapping timing optimization module is used to take the historical back corona intensity topology sequence, the predicted back corona intensity topology sequence, and the preset back corona intensity upper limit as inputs, take the rapping safety limit as constraints, and take minimizing the impact of secondary dust and minimizing the amount of premature rapping as optimization objectives to perform rapping sequence optimization and solve the optimal rapping timing scheme. The vibration execution control module is used to trigger the vibration execution devices of each corresponding partition in chronological order according to the optimal vibration timing scheme.

[0007] This application provides one or more technical solutions, which have at least the following technical effects or advantages: This application provides a method and system for staggered rapping timing control to suppress back corona. First, the dust collection electric field space is divided into multiple independent rapping control zones. This allows for individual control of rapping timing based on the back corona development state of different zones, avoiding a one-size-fits-all approach. Second, historical back corona intensity data is used to predict the back corona development trend within the rapping control cycle, providing accurate prediction data for rapping timing optimization. This allows for advance rapping based on the back corona development, avoiding the delayed problem of rapping only after the back corona has become severe. Finally, using a safe limit for back corona intensity as a constraint, and minimizing the impact of secondary dust and the amount of premature rapping as optimization objectives, a staggered rapping timing sequence is obtained. This staggered rapping avoids the large amount of secondary dust caused by simultaneous rapping in multiple zones, avoids unnecessary premature rapping, reduces wear on the rapping mechanism, and ensures the back corona suppression effect, maintaining the stable and efficient operation of the electrostatic precipitator and meeting stringent dust emission requirements.

[0008] Through the above technical solution, this application can effectively solve the problems of large secondary dust generation, untimely back corona suppression, and high wear of the rapping mechanism in the existing fixed unified rapping strategy. It can adapt to the increasingly stringent environmental emission standards and improve the operational stability and dust removal efficiency of electrostatic precipitators. It can be widely used in the operation control of electrostatic precipitators in various scenarios such as coal-fired power plants and industrial kilns. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a flowchart illustrating the rapping timing staggered peak control method for suppressing back corona provided in an embodiment of this application; Figure 2 This is a flowchart illustrating the process of determining multiple qualified rapping timing schemes in the rapping timing staggered peak control method for suppressing back corona provided in the embodiments of this application; Figure 3 This is a schematic diagram of the structure of the rapping timing misalignment control system for suppressing back corona provided in the embodiments of this application.

[0011] The components represented by each number in the attached diagram are explained below: Partition configuration module 11, back corona intensity assessment module 12, back corona intensity prediction module 13, rapping timing optimization module 14, and rapping execution control module 15. Detailed Implementation

[0012] This application provides a method and system for controlling the timing of rapping to suppress back corona, which addresses the technical problem in the prior art where unreasonable rapping timing leads to large secondary dust emissions and poor back corona suppression.

[0013] Example 1, as Figure 1 As shown, embodiments of this application provide a method for controlling the timing of rapping to suppress back corona, including: S10: The electrostatic precipitator's collecting electric field is spatially divided into several independent rapping control zones, and each zone is independently equipped with a rapping actuator and an electrical parameter monitoring unit. In this embodiment, the dust collection electric field is divided into multiple independent regions along the flue gas flow direction or the electric field height direction. Each region can be controlled to start and stop rapping independently without interfering with each other, which facilitates staggered rapping. That is, the entire electric field is divided into several groups, and each group is rapped at different times to avoid large-area dust generation occurring at the same time. The areas that are not rapped continue to maintain high-pressure dust collection to maintain the overall purification effect. The rapping timing is adjusted according to the back corona development in different regions to avoid the problem that traditional uniform rapping cannot adapt to regional differences.

[0014] Secondly, each zone is independently equipped with a rapping actuator and an electrical parameter monitoring unit to obtain real-time changes in electrical parameters during the operation of each zone, capture the occurrence and development of back corona in each zone, and provide accurate data support for subsequent rapping timing optimization.

[0015] S20: Through the electrical parameter monitoring unit, the secondary voltage timing data and secondary current timing data of each of the rapping control zones during operation are collected in real time, and the back corona severity index is calculated to determine the historical back corona intensity sequence of each of the rapping control zones in the historical monitoring period, and to construct the historical back corona intensity topology sequence of the dust collection electric field. In this embodiment, when back corona occurs, it causes a change in the discharge state within the electric field, specifically manifested as a decrease in secondary voltage and abnormal fluctuations in secondary current. Therefore, the severity index of back corona is calculated using the time-series data of secondary voltage and secondary current. The severity index of back corona in the current partition is stored in the corresponding location, and finally, a topological sequence of back corona intensity containing all partitions and the entire historical monitoring period can be obtained, which fully preserves the variation characteristics of back corona intensity in the spatial and temporal dimensions.

[0016] Specifically, real-time acquisition of secondary voltage and secondary current time-series data for each of the rapping control zones during operation is performed, and back corona severity index is calculated. The historical back corona intensity sequence for each of the rapping control zones within the historical monitoring period is determined, and a historical back corona intensity topology sequence of the dust collection electric field is constructed, including: According to the preset sampling frequency, the secondary voltage and secondary current values ​​of each vibration control zone are synchronously collected through the electrical parameter monitoring unit to obtain the historical secondary voltage sequence and historical secondary current sequence of each vibration control zone in the historical monitoring period. Based on the historical secondary voltage sequence and historical secondary current sequence of each rapping control zone, the volt-ampere characteristic slope and flashover frequency at each sampling moment are calculated as indicators of the severity of back corona. According to the preset back corona intensity grading mapping rule, the indicators of the severity of back corona are converted into the back corona intensity corresponding to the current sampling moment, and the historical back corona intensity sequence of each rapping control zone in the historical monitoring period is generated. Based on the position coordinates of each of the aforementioned rapping control zones, the corresponding historical back corona intensity sequences are combined and arranged to generate a historical back corona intensity topology sequence for the dust collection electric field.

[0017] In this embodiment, firstly, the electrical parameter monitoring unit synchronously collects the secondary voltage and secondary current data of each rapping control zone at a preset sampling frequency, such as every 10ms interval, to obtain the historical secondary voltage sequence and historical secondary current sequence of each rapping control zone in the historical monitoring period, ensuring that the data can reflect the development and changes of back corona in a timely manner, and avoiding the omission of the development process of back corona due to excessively large data sampling intervals.

[0018] Secondly, the severity of back corona is characterized by combining the slope of the current-voltage characteristic and the flashover frequency, two core indicators: after back corona occurs, the equivalent resistance inside the electric field will change, the slope of the current-voltage characteristic will change accordingly, and the flashover frequency will also increase significantly with the increase of the severity of back corona. Therefore, the current-voltage characteristic slope and the flashover frequency are combined to characterize the actual development state of back corona.

[0019] For example, the slope of the volt-ampere characteristic and the flashover frequency at each sampling moment can be calculated by using the sliding window method. Secondary voltage and secondary current data within a fixed time period before the current sampling moment are extracted, and the extracted voltage and current data are linearly fitted. The slope of the fitted line is the slope of the volt-ampere characteristic at that sampling moment. The number of flashovers within the fixed time period before the current sampling moment is counted, and the flashover frequency at that sampling moment is obtained by dividing it by the fixed time period.

[0020] Furthermore, by using a preset hierarchical mapping rule, continuous index values ​​are mapped to corresponding levels of back corona intensity, facilitating subsequent prediction and constraint judgment. The preset hierarchical mapping rule divides the slope of the volt-ampere characteristic into several numerical intervals, and also divides the flashover frequency into several numerical intervals. Different interval combinations correspond to different levels of back corona intensity. The higher the level, the more severe the back corona. Finally, the back corona intensity corresponding to each sampling moment is obtained, and the historical back corona intensity sequence of each rapping control zone within the historical monitoring period can be generated.

[0021] Finally, by combining and arranging the historical back corona intensity sequences corresponding to each zone according to the actual spatial coordinates of each rapping control zone, a topological sequence of historical back corona intensity of the dust collection electric field that simultaneously contains spatial location information and temporal change information can be obtained.

[0022] Specifically, the steps for constructing the anti-corona intensity grading mapping rule include: The reference values ​​of the volt-ampere characteristic slope and flashover frequency of the electrostatic precipitator are obtained when it is operating normally under calibrated conditions. The reference values ​​of the volt-ampere characteristic slope and flashover frequency are rolled over according to a preset update cycle based on the real-time operating data of the electrostatic precipitator under non-back corona conditions. Based on the reference value of the current-voltage characteristic slope and the reference value of the flashover frequency, multiple current-voltage characteristic slope thresholds and multiple flashover frequency thresholds are set respectively, wherein the multiple current-voltage characteristic slope thresholds decrease sequentially along the direction of increasing back corona intensity, and the multiple flashover frequency thresholds increase sequentially along the direction of increasing back corona intensity. Based on the multiple volt-ampere characteristic slope thresholds and multiple flashover frequency thresholds, multiple continuous and non-overlapping back corona intensity level intervals are divided, and a unique back corona intensity level value is assigned to each back corona intensity level interval, forming a mapping relationship from the volt-ampere characteristic slope, the flashover frequency to the back corona intensity level value.

[0023] In this embodiment of the application, a classification mapping rule for back corona intensity is constructed. First, the reference values ​​of the volt-ampere characteristic slope and flashover frequency of the electrostatic precipitator under normal operation under calibrated conditions are obtained. The reference values ​​are the values ​​measured when the equipment is running stably without significant back corona, representing the basic state when the back corona intensity is 0. The reference values ​​of the volt-ampere characteristic slope and flashover frequency are updated using a rolling window adaptive update strategy under the condition of no back corona.

[0024] Among them, the non-corona discharge condition includes normal voltage and current characteristics, normal flashover frequency, recovery period after rapping, and stable operation. Specifically, normal voltage and current characteristics means that the ratio of the current secondary voltage to the secondary current is within the historical normal range, and the slope of the volt-ampere characteristic has not shown an abnormal decrease; normal flashover frequency means that the current flashover frequency is lower than the preset low-frequency threshold, eliminating the state of frequent flashover; recovery period after rapping means that the time since the end of the last rapping exceeds the preset recovery time, avoiding dust interference caused by rapping; stable operation means that key process parameters such as flue gas temperature, flue gas flow rate, and inlet dust concentration fluctuate within the preset stable range.

[0025] Furthermore, based on the rolling window adaptive update strategy, under the premise of ensuring no back corona, the benchmark value is dynamically corrected using real-time operating data. According to the preset update cycle, such as triggering the benchmark value correction once every 24 hours of operation, during the correction process, only the operating data currently in the state without back corona is selected for updating, avoiding the mixing of abnormal data in the back corona state into the benchmark value. The moving average method is used to weight and fuse the historical benchmark value with the current measured value to generate the updated benchmark value, and an update rate limit is set, such as the single update correction amount not exceeding ±1% of the original benchmark value, to avoid large changes in the benchmark value.

[0026] The moving average method refers to selecting the slope of the measured volt-ampere characteristic and the flashover frequency of all the measured values ​​that meet the requirements of the non-reverse corona condition within the current update cycle, calculating the measured average value, multiplying the historical benchmark value by the weighting coefficient α, and adding the measured average value multiplied by the weighting coefficient (1-α) to obtain the benchmark value after the current update. The weighting coefficient α is a preset smoothing coefficient, and its value range is usually set to 0.6~0.8.

[0027] Subsequently, multiple thresholds are set based on the reference values ​​of the current-voltage characteristic slope and the flashover frequency. The current-voltage characteristic slope gradually decreases as the back corona develops, so its threshold decreases sequentially along the direction of increasing back corona intensity; the flashover frequency gradually increases as the back corona develops, so its threshold increases sequentially along the direction of increasing back corona intensity.

[0028] For example, the slope benchmark value of the volt-ampere characteristic can be set as k0, and three thresholds k1, k2, and k3 can be set to satisfy k0 > k1 > k2 > k3, corresponding to four slope intervals; three flashover frequency thresholds f1, f2, and f3 can be set to satisfy f1 < f2 < f3, corresponding to four frequency intervals. Different slope intervals and frequency intervals can be combined to obtain 4×4, a total of 16 non-overlapping interval combinations. Each combination is assigned a unique back corona intensity level value from 0 to 15. The larger the level value, the more severe the back corona. This completes the construction of the hierarchical mapping rule.

[0029] Finally, after dividing multiple non-overlapping back corona intensity level intervals by different threshold combinations, a unique back corona intensity level value is assigned to each back corona intensity level interval, thus completing the construction of the hierarchical mapping rule and realizing the standardized quantification of the severity of back corona. The judgment by combining two indicators can more accurately reflect the actual development degree of back corona than a single indicator, avoiding misjudgment caused by the interference of operating condition fluctuations of a single indicator, and improving the accuracy of back corona intensity assessment.

[0030] S30: Based on the historical back corona intensity topology sequence, predict the back corona intensity sequence of each of the rapping control zones within the preset rapping control cycle, and construct the predicted back corona intensity topology sequence of the dust collection electric field. In this embodiment, the spatial location and temporal variation information contained in the historical back corona intensity topology sequence are used to predict the back corona intensity changes in each rapping control zone within future rapping control cycles through a time series prediction model combined with spatial correlation information. Since the back corona development in different zones is not completely independent, the ash accumulation state and back corona development in adjacent zones will influence each other. Therefore, combining spatial correlation features with temporal trend changes in prediction yields prediction results that better reflect the actual development situation.

[0031] Specifically, step S30 in the method includes: Using the historical monitoring cycle as the input time window length and the rapping control cycle as the output time window length, sample data is extracted from the historical operation monitoring records of the dust collection electric field. For each set of sample data, the anti-corona intensity topology sequence that satisfies the input time window length within a continuous time period is selected as the sample input sequence, and the anti-corona intensity topology sequence that is temporally adjacent to the sample input sequence and satisfies the output time window length is selected as the sample output sequence. The sample input sequence is paired with the sample output sequence to construct a sample dataset for training the prediction model; The Long Short-Term Memory network was trained to converge using the aforementioned sample dataset to generate an anti-corona intensity predictor. The historical back corona intensity topology sequence is input into the back corona intensity predictor, and the predicted back corona intensity topology sequence of the dust collection electric field during the rapping control cycle is output.

[0032] In this embodiment of the application, firstly, the historical monitoring period is used as the input time window length and the rapping control period is used as the output time window length. For example, the historical monitoring period length is 2 hours and the rapping control period length is the next 30 minutes. A large number of samples are extracted from the historical monitoring data of the electrostatic precipitator during long-term operation to ensure that the model can learn the development law of back corona under different working conditions.

[0033] Secondly, for each group of extracted samples, the anti-corona intensity topology sequence that meets the input time window length within a continuous time period is first taken as input, and then the anti-corona intensity topology sequence that is temporally adjacent to the input sequence time is taken as output. All samples are paired in this way to obtain the complete training dataset.

[0034] Subsequently, the dataset is used to train the preset long short-term memory network until the network loss function converges, resulting in a trained anti-corona intensity predictor. Finally, the currently obtained historical anti-corona intensity topology sequence is input into the trained anti-corona intensity predictor, which can directly output the predicted anti-corona intensity topology sequence of the entire dust collection electric field in the future rapping control cycle, thus completing the prediction of the anti-corona development trend and providing an accurate basis for subsequent rapping timing optimization.

[0035] For example, a corona intensity predictor is trained based on a Long Short-Term Memory (LSTM) network. First, an LSTM network model is constructed using a two-layer LSTM structure. The first LSTM layer has 64 hidden state units and returns the output for each time step. The second LSTM layer has 32 hidden state units. The output layer uses a temporally distributed fully connected layer to map the hidden state at each time step to a corona intensity prediction value equal to the number of partitions. The Adam optimizer is used, with an initial learning rate of 0.001, a mean squared error loss function, and a mean absolute error as the evaluation metric. The batch size is set to 32, the maximum training epochs are 100, and an early stopping mechanism is introduced: training is terminated early if the validation set loss does not decrease for 10 consecutive epochs.

[0036] During training, the sample input sequence of each batch is fed into the LSTM network for forward propagation. The mean squared error loss between the predicted output and the true output is calculated. The network weight parameters and bias term parameters are updated by backpropagation using the Adam optimizer. After each round of training, the mean absolute error loss is evaluated on the validation set until the model converges. After training converges, the network parameters are saved to obtain the anti-corona intensity predictor.

[0037] S40: Using the historical back corona intensity topology sequence, the predicted back corona intensity topology sequence, and the preset back corona intensity upper limit as inputs, with the rapping safety limit as a constraint, and with minimizing the impact of secondary dust and minimizing the amount of premature rapping as optimization objectives, perform rapping sequence optimization to solve for the optimal rapping timing scheme. In this embodiment, based on the predicted changes in back corona intensity of each zone within the future rapping control cycle and combined with the back corona development pattern obtained from historical monitoring, the rapping time of all zones to be rapped is arranged and optimized under the premise of satisfying rapping safety constraints, to obtain the globally optimal staggered rapping timing arrangement.

[0038] Specifically, the rapping safety limit constraint means that any two adjacent rapping control zones are not allowed to rap simultaneously within the same time period to prevent the simultaneous rapping of adjacent zones from generating superimposed dust and significantly increasing the dust concentration at the outlet. At the same time, it requires that the rapping interval of any zone is not less than the preset minimum rapping interval to avoid repeated rapping in a short period of time, which would increase unnecessary mechanical wear. It also requires that all zones whose predicted back corona intensity exceeds the preset upper limit of back corona intensity must be rapped within the current rapping control cycle to prevent the long-term development of back corona from significantly reducing dust collection efficiency.

[0039] Specifically, step S40 in the method includes: Obtain a preset upper limit value for back corona intensity, and correct the upper limit value for back corona intensity downward according to a preset redundancy range to generate a limit value for back corona intensity. Based on the preset back corona intensity execution threshold, the historical back corona intensity topology sequence is filtered, and the vibration control partitions with historical back corona intensity greater than the back corona intensity execution threshold are selected as vibration partitions to be vibrated, thus determining the set of vibration partitions to be vibrated. Based on the predicted back corona intensity sequence corresponding to each of the partitions to be vibrated, calculate the latest allowed vibration triggering time under the premise that the predicted back corona intensity is always less than the back corona intensity limit value, and use it as the vibration cutoff time. Based on the predicted back corona intensity topology sequence and the set of partitions to be rapped, with the rapping trigger time of all partitions to be rapped not later than their respective rapping cutoff time as a hard constraint, and with the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, the rapping sequence optimization is performed to solve for the optimal rapping timing scheme.

[0040] In this embodiment, the back corona intensity is a dynamic parameter that changes continuously over time. If the preset upper limit of the back corona intensity is directly used as the critical point for triggering the rapping, the back corona phenomenon will have already occurred when the back corona intensity actually reaches the upper limit. At this time, the secondary current has increased sharply and the dust removal efficiency has been substantially damaged. Therefore, this application obtains a back corona intensity warning limit value that is lower than the actual upper limit by setting a redundant range and correcting it downward, so that the rapping is triggered in advance before the back corona intensity reaches the dangerous upper limit, and the dust removal process is completed before the back corona really deteriorates.

[0041] Specifically, firstly, a preset upper limit value for back corona intensity is obtained. Then, the upper limit value for back corona intensity is corrected downward according to a preset redundancy interval to obtain the back corona intensity limit value. The preset redundancy interval adjusts the width of the interval according to the actual operational stability requirements. The larger the redundancy interval, the lower the corrected back corona intensity limit value, the earlier the rapping triggering time, and the better the prevention of back corona. However, it will also increase the total number of early rappings, increasing the probability of secondary dust emission. Therefore, the redundancy interval can be set according to the on-site outlet dust emission control requirements. For example, in this application, it is set to 10%~15% of the upper limit value to balance the early prevention effect and the risk of secondary dust emission.

[0042] Furthermore, the preset redundancy interval is not a fixed value, but is dynamically determined based on the evolution rate of the back corona intensity of each rapping control zone. Specifically, when the average evolution rate of the back corona intensity of a zone is higher than the preset reference rate, the redundancy interval of that zone is increased, so that the back corona intensity threshold value is reduced accordingly, thereby advancing the rapping trigger time; when the evolution rate is lower than the reference rate, the redundancy interval is reduced, and the rapping trigger time is appropriately delayed. The evolution rate of each zone is calculated independently, and the redundancy interval is adjusted independently, realizing adaptive staggered rapping of each zone.

[0043] Secondly, based on the preset back corona intensity execution threshold, the partitions that have already shown a certain degree of back corona are screened, and the partitions with historical back corona intensity greater than the execution threshold are included in the set of partitions to be rapped. The partitions whose back corona intensity is still within the safe range and do not need to be rapped this time are filtered out. The preset back corona intensity is a safe threshold below the back corona intensity limit value. Only the partitions that do have back corona risk and need to be cleaned are screened, reducing the amount of calculation in the optimization process. For example, it can be set to 80% of the back corona intensity limit value, which avoids missing the partitions that need to be cleaned and also avoids including too many partitions that do not need to be cleaned, thus reducing the calculation burden.

[0044] Secondly, for each selected zone to be rapped, the predicted back corona intensity change sequence of that zone is checked one by one from the prediction start time. On the basis that the predicted back corona intensity is always less than the back corona intensity limit value, the latest allowed rapping trigger time, i.e. the rapping cutoff time, is found. As long as the rapping trigger time is not later than the cutoff time, it can be guaranteed that the dust removal is completed before the back corona intensity reaches the warning value, thus avoiding the back corona exceeding the standard.

[0045] Finally, based on the predicted back corona intensity topology sequence and the set of partitions to be rapped obtained above, the following hard constraints are set: the rapping trigger time of all partitions to be rapped is no later than their respective rapping cutoff time, adjacent partitions are not rapped at the same time, and the rapping interval meets the minimum requirement. With the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, global optimization of the rapping sequence is carried out, and the optimal rapping sequence scheme that satisfies all safety constraints can be obtained.

[0046] Specifically, based on the predicted back corona intensity topology sequence and the set of partitions to be rapped, with the rapping trigger time of all partitions to be rapped not later than their respective rapping cutoff time as a hard constraint, and with the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, the rapping sequence optimization is performed to solve for the optimal rapping timing scheme, including: The maximum number of simultaneous vibration triggers is obtained, and the number of vibration zones to be vibrated at the same time is less than or equal to the maximum number of simultaneous vibration triggers is used as a cooperative vibration constraint condition. Based on the set of partitions to be vibrated, under the premise of satisfying the cooperative vibration constraint, the vibration order of each partition to be vibrated is randomly arranged and the scheme is enumerated to generate several initial vibration timing schemes. Using the constraint that the vibration trigger time of all the partitions to be vibrated is no later than their respective vibration cutoff time as a hard constraint, the several initial vibration timing schemes are screened to determine several qualified vibration timing schemes. Based on the predicted back corona intensity topology sequence, with the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, the quality of the multiple qualified rapping timing schemes is evaluated, and the qualified rapping timing scheme corresponding to the scheme with the largest scheme quality coefficient is selected as the optimal rapping timing scheme.

[0047] In the embodiments of this application, such as Figure 2 As shown, firstly, the maximum number of simultaneous rapping triggers is set. This parameter is determined based on the allowable instantaneous dust concentration at the electrostatic precipitator outlet. The higher the allowable concentration, the larger the maximum number of simultaneous rapping can be set, and vice versa. This parameter limits the total number of zones that can be rapped simultaneously from a global perspective, avoiding large-scale secondary dust generation caused by multiple zones rapping simultaneously. For example, for an electric field, if the maximum number of simultaneous rapping is 6, then the number of zones that can be rapped simultaneously is less than or equal to 6 and greater than or equal to 1, that is, at least one zone is arranged for rapping to ensure that the dust removal process continues without long-term stagnation.

[0048] Secondly, based on the set of partitions to be vibrated, and using the collaborative vibration constraint as a basis, multiple initial vibration timing schemes are randomly generated. Then, with the vibration trigger time of all partitions to be vibrated not being later than the deadline time as a screening condition, illegal schemes that cannot meet the hard constraint requirements are eliminated, resulting in multiple qualified schemes.

[0049] Finally, for all qualified schemes, based on the predicted back corona intensity topology sequence, the optimization objective function value corresponding to each scheme is calculated. The corresponding scheme quality coefficient is obtained through objective function transformation. The scheme with the largest coefficient is selected as the optimal rapping timing scheme for the final output. This can minimize the impact of secondary dust while satisfying all safety constraints, and delay the rapping timing as much as possible to reduce unnecessary early dust removal, thus balancing the needs of back corona prevention and outlet dust stability control.

[0050] Furthermore, based on the predicted back corona intensity topology sequence, and with the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, the quality of the multiple qualified rapping timing schemes is evaluated, including: For each qualified rapping timing scheme, based on the predicted back corona intensity topology sequence, the rapping trigger time of each partition to be rapped in the current scheme and the predicted back corona intensity value corresponding to the current time are extracted. Based on the rapping trigger time of each zone to be rapped, calculate the quantification value of the secondary dust impact of the current qualified rapping sequence scheme; Based on the predicted back corona intensity value of each zone to be vibrated and the back corona intensity limit value, calculate the advance vibration quantization value of the current qualified vibration timing scheme; After dimensionless processing of the quantified values ​​of the secondary dust impact and the quantified values ​​of the early rapping, the reciprocals of the two are weighted and summed to obtain the scheme quality coefficient of the current qualified rapping timing scheme.

[0051] In this embodiment, firstly, for each zone to be rapped, the rapping trigger time arranged in the qualified rapping timing scheme is extracted, and then the predicted back corona intensity corresponding to that time is obtained by combining the topological sequence. The higher the current back corona intensity, the thicker the dust accumulation in the zone. Then, based on the rapping trigger time of each zone to be rapped, the maximum number of rappings, the average number of rappings, and the coefficient of variation of the number of rappings are obtained, and the secondary dust impact quantification value of the current qualified rapping timing scheme is obtained by weighted calculation.

[0052] Secondly, for each zone to be vibrated, the difference between the back corona intensity threshold and the predicted back corona intensity at the vibration trigger time is calculated. The larger the difference, the earlier the vibration is triggered and the higher the degree of early vibration. Therefore, by summing up the differences of all zones to be vibrated, the early vibration quantification value of the entire scheme can be obtained.

[0053] Finally, the quantitative values ​​of the secondary dust impact and the early rapping were subjected to dimensionless normalization. The reciprocals of the two normalization results were taken and then summed according to preset weights to obtain the quality coefficient of the scheme. The dimensionless processing eliminated the difference in magnitude between the two quantitative indicators. After reciprocal weighting, the smaller the secondary dust impact and the amount of early rapping, the larger the quality coefficient obtained.

[0054] For example, dimensionless normalization of the quantitative values ​​of secondary dust impact and early rapping can be achieved through maximum and minimum value normalization. The two indicators of all qualified solutions are mapped to the 0-1 range respectively. The preset weights can be adjusted according to the on-site emission control requirements. If the stability of the outlet dust concentration is required to be high, the weight of the secondary dust impact item is increased. If more emphasis is placed on reducing the mechanical loss of rapping, the weight of the early rapping amount item is increased.

[0055] Specifically, based on the rapping trigger time of each zone to be rapped, the secondary dust impact quantification value of the current qualified rapping sequence scheme is calculated, including: The vibration control cycle is divided into multiple consecutive discrete time slices; The number of partitions to be vibrated in the vibrating state within each discrete time slice is counted, a time slice vibrating number sequence is generated, and the maximum vibrating number of the time slice vibrating number sequence is determined. Calculate the mean number of rappings and the coefficient of variation of the number of rappings in the time-slice rapping sequence; The secondary dust impact quantification value of the current qualified rapping sequence scheme is determined based on the maximum number of rappings, the average number of rappings, and the coefficient of variation of the number of rappings. The secondary dust impact quantification value is positively correlated with the maximum number of rappings, the average number of rappings, and the coefficient of variation of the number of rappings.

[0056] In this embodiment, the entire vibration control cycle is first divided into discrete time slices of equal length. The length of the time slice is matched with the duration of a single vibration, ensuring that the vibration state remains stable within each time slice. For example, if the duration of a single vibration is 3 seconds, the length of the time slice is also set to 3 seconds, which facilitates accurate counting of the number of partitions vibrating simultaneously at each moment.

[0057] Secondly, the number of partitions to be rapped in the rapping state within each discrete time slice is counted one by one to obtain a complete time slice rapping number sequence. The maximum rapping number in the sequence is then extracted. The maximum rapping number reflects the instantaneous maximum dust intensity. The larger the value, the higher the instantaneous dust concentration and the greater the impact on emissions.

[0058] Next, the mean number of rapping events in all time slots of the sequence is calculated to reflect the overall average dust level of the scheme. At the same time, the coefficient of variation of the number of rapping events is calculated to reflect the uniformity of the distribution of the number of rapping events in different time slots. The larger the coefficient of variation, the more concentrated the rapping arrangement is, and the more likely there will be a sudden increase in instantaneous concentration. For example, the coefficient of variation can be calculated by dividing the standard deviation by the mean number of rapping events, which can reflect the uniformity of the rapping arrangement.

[0059] Finally, the maximum number of rapping beats, the average number of rapping beats, and the coefficient of variation of the number of rapping beats are weighted and summed according to preset weights to obtain the quantitative value of the secondary dust impact of the current scheme. The preset weight coefficients are obtained through experimental calibration. Among them, the maximum number of rapping beats has the highest weight because the maximum number of simultaneous rapping beats directly determines the peak value of instantaneous outlet dust concentration and has the greatest impact on emission compliance. The average number of rapping beats and the coefficient of variation of the number of rapping beats correspond to the average dust level and distribution uniformity for auxiliary evaluation, respectively. For example, the weight of the maximum number of rapping beats can be 0.5, and the average number of rapping beats and the coefficient of variation of the number of rapping beats can be 0.3 and 0.2, respectively. All three parameters are positively correlated with the secondary dust impact, and increasing any one of the parameters will lead to an increase in the total quantitative value.

[0060] Further, based on the predicted back corona intensity value of each zone to be rapped and the back corona intensity limit value, the advance rapping quantization value of the current qualified rapping timing scheme is calculated, including: For each of the partitions to be rapped, the difference between the back corona intensity limit value and the predicted back corona intensity value of the current partition at the rapping trigger time is calculated as the advance rapping sub-quantity of the current partition. The advance rapping sub-quantities of all the sections to be rapped are summed, and the summation result is used as the advance rapping quantization value of the current qualified rapping timing scheme.

[0061] In this embodiment, firstly, for each zone to be rapped, the back corona intensity limit value is a pre-corrected safety warning value. The further the predicted back corona intensity value of the zone is from the limit value at the rapping trigger time, the earlier the rapping is triggered, the greater the advance amount of the dust removal operation, and the more unfavorable it is to the equipment and emission control. Therefore, the predicted back corona intensity value at the rapping trigger time is subtracted from the back corona intensity limit value to obtain the advance rapping amount of the zone. The greater the difference, the greater the advance rapping amount and the higher the degree of advance.

[0062] Secondly, the advance rapping sub-quantities of all selected rapping zones are enumerated one by one to obtain the total advance rapping quantification value of the entire rapping timing scheme. The larger the accumulated result, the higher the overall advance degree of rapping in the entire scheme, the more unnecessary advance dust removal times, and the worse the scheme quality.

[0063] S50: According to the optimal vibration timing scheme, the vibration execution devices of each corresponding partition are triggered sequentially in time order.

[0064] In this embodiment of the application, after obtaining the optimal rapping timing scheme that satisfies all constraints, trigger commands can be sent sequentially to the rapping execution devices corresponding to each partition to be rapped according to the rapping trigger times arranged in the scheme, and the dust removal operation of the corresponding partition can be completed in sequence.

[0065] The entire rapping timing scheme is a globally optimal arrangement obtained for the current calculation cycle. During execution, no temporary adjustments are needed; rapping simply needs to be triggered at the predetermined time. This ensures that the corresponding zone is cleaned before the back corona intensity reaches the warning threshold. Simultaneously, it maximizes the staggering of rapping times for multiple zones, avoiding sudden increases in outlet dust concentration caused by concentrated rapping, and also delays rapping timing as much as possible, reducing unnecessary premature rapping and balancing the needs of back corona suppression and dust emission control. At the start of the next rapping control cycle, the operating data for each zone is recollected, and the process of predicting back corona intensity, selecting zones to be rapped, and optimizing the rapping timing is repeated to update the optimal rapping timing scheme for the next cycle, achieving dynamic closed-loop control.

[0066] In summary, compared with existing technologies, this application, by staggering the timing of rapping, disperses the rapping of zones with back corona risk at different times, avoiding simultaneous rapping of adjacent zones or concentrated rapping of a large number of zones. This not only completes dust removal before the back corona intensity exceeds the standard, thus suppressing the development of back corona in advance, but also minimizes the secondary dust impact caused by concentrated rapping, avoiding a sudden increase in outlet dust concentration that leads to excessive emissions. At the same time, by optimizing the timing of rapping as much as possible, unnecessary early rapping is reduced, and mechanical wear of the rapping is decreased. This approach balances the back corona suppression effect, emission stability, and equipment lifespan, solving the problem of the contradiction between back corona prevention and secondary dust control in existing rapping control logic.

[0067] In summary, the embodiments of this application have at least the following technical effects: This application provides a method for staggered rapping timing to suppress back corona. First, the dust collection electric field space is divided into multiple independent rapping control zones. This allows for individual control of rapping timing based on the back corona development state of different zones, avoiding a one-size-fits-all approach. Second, historical back corona intensity data is used to predict the back corona development trend within the rapping control cycle, providing accurate prediction data for rapping timing optimization. This allows for advance rapping based on back corona development, avoiding the lag problem of rapping only after the back corona has become severe. Finally, using a safe limit for back corona intensity as a constraint, and minimizing the impact of secondary dust and the amount of premature rapping as optimization objectives, a staggered rapping timing sequence is obtained. This staggered rapping avoids the large amount of secondary dust caused by simultaneous rapping in multiple zones, avoids unnecessary premature rapping, reduces wear on the rapping mechanism, and ensures the back corona suppression effect, maintaining the stable and efficient operation of the electrostatic precipitator and meeting stringent dust emission requirements.

[0068] Through the above technical solution, this application can effectively solve the problems of large secondary dust generation, untimely back corona suppression, and high wear of the rapping mechanism in the existing fixed unified rapping strategy. It can adapt to the increasingly stringent environmental emission standards and improve the operational stability and dust removal efficiency of electrostatic precipitators. It can be widely used in the operation control of electrostatic precipitators in various scenarios such as coal-fired power plants and industrial kilns.

[0069] Example 2, as Figure 3 As shown, based on the same inventive concept as the vibration timing misalignment control method for suppressing back corona provided in Embodiment 1, this application also provides a vibration timing misalignment control system for suppressing back corona, including: The partition configuration module 11 is used to divide the dust collection electric field of the electrostatic precipitator into several independent rapping control partitions in space, and to equip each partition with an independent rapping actuator and an electrical parameter monitoring unit. The back corona intensity assessment module 12 is used to collect the secondary voltage time sequence data and secondary current time sequence data of each of the rapping control zones in real time during operation through the electrical parameter monitoring unit, calculate the back corona severity index, determine the historical back corona intensity sequence of each of the rapping control zones in the historical monitoring period, and construct the historical back corona intensity topology sequence of the dust collection electric field. The back corona intensity prediction module 13 is used to predict and obtain the predicted back corona intensity sequence of each of the rapping control zones within a preset rapping control cycle based on the historical back corona intensity topology sequence, and to construct the predicted back corona intensity topology sequence of the dust collection electric field. The rapping timing optimization module 14 is used to take the historical back corona intensity topology sequence, the predicted back corona intensity topology sequence and the preset back corona intensity upper limit as input, take the rapping safety limit as constraint, and take minimizing the impact of secondary dust and minimizing the amount of premature rapping as optimization objectives to perform rapping sequence optimization and solve the optimal rapping timing scheme. The vibration execution control module 15 is used to trigger the vibration execution devices of each corresponding partition in sequence according to the optimal vibration timing scheme.

[0070] Further, in one embodiment, the secondary voltage and secondary current time-series data of each of the rapping control zones are collected in real time during operation, and the back corona severity index is calculated to determine the historical back corona intensity sequence of each of the rapping control zones within the historical monitoring period, constructing a historical back corona intensity topology sequence of the dust collection electric field, including: According to the preset sampling frequency, the secondary voltage and secondary current values ​​of each vibration control zone are synchronously collected through the electrical parameter monitoring unit to obtain the historical secondary voltage sequence and historical secondary current sequence of each vibration control zone in the historical monitoring period. Based on the historical secondary voltage sequence and historical secondary current sequence of each rapping control zone, the volt-ampere characteristic slope and flashover frequency at each sampling moment are calculated as indicators of the severity of back corona. According to the preset back corona intensity grading mapping rule, the indicators of the severity of back corona are converted into the back corona intensity corresponding to the current sampling moment, and the historical back corona intensity sequence of each rapping control zone in the historical monitoring period is generated. Based on the position coordinates of each of the aforementioned rapping control zones, the corresponding historical back corona intensity sequences are combined and arranged to generate a historical back corona intensity topology sequence for the dust collection electric field.

[0071] Furthermore, the construction steps of the anti-corona intensity grading mapping rule include: The reference values ​​of the volt-ampere characteristic slope and flashover frequency of the electrostatic precipitator are obtained when it is operating normally under calibrated conditions. The reference values ​​of the volt-ampere characteristic slope and flashover frequency are rolled over according to a preset update cycle based on the real-time operating data of the electrostatic precipitator under non-back corona conditions. Based on the reference value of the current-voltage characteristic slope and the reference value of the flashover frequency, multiple current-voltage characteristic slope thresholds and multiple flashover frequency thresholds are set respectively, wherein the multiple current-voltage characteristic slope thresholds decrease sequentially along the direction of increasing back corona intensity, and the multiple flashover frequency thresholds increase sequentially along the direction of increasing back corona intensity. Based on the multiple volt-ampere characteristic slope thresholds and multiple flashover frequency thresholds, multiple continuous and non-overlapping back corona intensity level intervals are divided, and a unique back corona intensity level value is assigned to each back corona intensity level interval, forming a mapping relationship from the volt-ampere characteristic slope, the flashover frequency to the back corona intensity level value.

[0072] In one embodiment, the back corona intensity prediction module 13 is specifically used for: Using the historical monitoring cycle as the input time window length and the rapping control cycle as the output time window length, sample data is extracted from the historical operation monitoring records of the dust collection electric field. For each set of sample data, the anti-corona intensity topology sequence that satisfies the input time window length within a continuous time period is selected as the sample input sequence, and the anti-corona intensity topology sequence that is temporally adjacent to the sample input sequence and satisfies the output time window length is selected as the sample output sequence. The sample input sequence is paired with the sample output sequence to construct a sample dataset for training the prediction model; The Long Short-Term Memory network was trained to converge using the aforementioned sample dataset to generate an anti-corona intensity predictor. The historical back corona intensity topology sequence is input into the back corona intensity predictor, and the predicted back corona intensity topology sequence of the dust collection electric field during the rapping control cycle is output.

[0073] In one embodiment, the vibration timing optimization module 14 is specifically used for: Obtain a preset upper limit value for back corona intensity, and correct the upper limit value for back corona intensity downward according to a preset redundancy range to generate a limit value for back corona intensity. Based on the preset back corona intensity execution threshold, the historical back corona intensity topology sequence is filtered, and the vibration control partitions with historical back corona intensity greater than the back corona intensity execution threshold are selected as vibration partitions to be vibrated, thus determining the set of vibration partitions to be vibrated. Based on the predicted back corona intensity sequence corresponding to each of the partitions to be vibrated, calculate the latest allowed vibration triggering time under the premise that the predicted back corona intensity is always less than the back corona intensity limit value, and use it as the vibration cutoff time. Based on the predicted back corona intensity topology sequence and the set of partitions to be rapped, with the rapping trigger time of all partitions to be rapped not later than their respective rapping cutoff time as a hard constraint, and with the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, the rapping sequence optimization is performed to solve for the optimal rapping timing scheme.

[0074] Furthermore, based on the predicted back corona intensity topology sequence and the set of partitions to be rapped, with the rapping trigger time of all partitions to be rapped not later than their respective rapping cutoff time as a hard constraint, and with the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, the rapping sequence optimization is performed to solve for the optimal rapping timing scheme, including: The maximum number of simultaneous vibration triggers is obtained, and the number of vibration zones to be vibrated at the same time is less than or equal to the maximum number of simultaneous vibration triggers is used as a cooperative vibration constraint condition. Based on the set of partitions to be vibrated, under the premise of satisfying the cooperative vibration constraint, the vibration order of each partition to be vibrated is randomly arranged and the scheme is enumerated to generate several initial vibration timing schemes. Using the constraint that the vibration trigger time of all the partitions to be vibrated is no later than their respective vibration cutoff time as a hard constraint, the several initial vibration timing schemes are screened to determine several qualified vibration timing schemes. Based on the predicted back corona intensity topology sequence, with the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, the quality of the multiple qualified rapping timing schemes is evaluated, and the qualified rapping timing scheme corresponding to the scheme with the largest scheme quality coefficient is selected as the optimal rapping timing scheme.

[0075] Furthermore, based on the predicted back corona intensity topology sequence, and with the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, the quality of the multiple qualified rapping timing schemes is evaluated, including: For each qualified rapping timing scheme, based on the predicted back corona intensity topology sequence, the rapping trigger time of each partition to be rapped in the current scheme and the predicted back corona intensity value corresponding to the current time are extracted. Based on the rapping trigger time of each zone to be rapped, calculate the quantification value of the secondary dust impact of the current qualified rapping sequence scheme; Based on the predicted back corona intensity value of each zone to be vibrated and the back corona intensity limit value, calculate the advance vibration quantization value of the current qualified vibration timing scheme; After dimensionless processing of the quantified values ​​of the secondary dust impact and the quantified values ​​of the early rapping, the reciprocals of the two are weighted and summed to obtain the scheme quality coefficient of the current qualified rapping timing scheme.

[0076] Furthermore, in one embodiment of the application, the secondary dust impact quantification value of the current qualified rapping timing scheme is calculated based on the rapping trigger time of each zone to be rapped, including: The vibration control cycle is divided into multiple consecutive discrete time slices; The number of partitions to be vibrated in the vibrating state within each discrete time slice is counted, a time slice vibrating number sequence is generated, and the maximum vibrating number of the time slice vibrating number sequence is determined. Calculate the mean number of rappings and the coefficient of variation of the number of rappings in the time-slice rapping sequence; The secondary dust impact quantification value of the current qualified rapping sequence scheme is determined based on the maximum number of rappings, the average number of rappings, and the coefficient of variation of the number of rappings. The secondary dust impact quantification value is positively correlated with the maximum number of rappings, the average number of rappings, and the coefficient of variation of the number of rappings.

[0077] Further, based on the predicted back corona intensity value of each zone to be rapped and the back corona intensity limit value, the advance rapping quantization value of the current qualified rapping timing scheme is calculated, including: For each of the partitions to be rapped, the difference between the back corona intensity limit value and the predicted back corona intensity value of the current partition at the rapping trigger time is calculated as the advance rapping sub-quantity of the current partition. The advance rapping sub-quantities of all the sections to be rapped are summed, and the summation result is used as the advance rapping quantization value of the current qualified rapping timing scheme.

[0078] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for inhibiting back corona by rapping timing de-peak regulation, characterized in that, The methods include: The electrostatic precipitator's collecting electric field is spatially divided into several independent rapping control zones, and each zone is independently equipped with a rapping actuator and an electrical parameter monitoring unit. The electrical parameter monitoring unit collects the secondary voltage and secondary current time sequence data of each rapping control zone in real time during operation, calculates the back corona severity index, determines the historical back corona intensity sequence of each rapping control zone in the historical monitoring period, and constructs the historical back corona intensity topology sequence of the dust collection electric field. Based on the historical back corona intensity topology sequence, the predicted back corona intensity sequence of each of the rapping control zones within the preset rapping control cycle is obtained, and the predicted back corona intensity topology sequence of the dust collection electric field is constructed. Using the historical back corona intensity topology sequence, the predicted back corona intensity topology sequence, and the preset back corona intensity upper limit as inputs, with the rapping safety limit as a constraint, and with the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, the rapping sequence optimization is performed to solve for the optimal rapping timing scheme. According to the optimal vibration timing scheme, the vibration execution devices of each corresponding zone are triggered sequentially in time order.

2. The method for suppressing back corona beat timing peak regulation according to claim 1, characterized in that, Real-time acquisition of secondary voltage and secondary current time-series data for each rapping control zone during operation, calculation of back corona severity index, determination of historical back corona intensity sequence for each rapping control zone within the historical monitoring period, and construction of historical back corona intensity topology sequence for the dust collection electric field, including: According to the preset sampling frequency, the secondary voltage and secondary current values ​​of each vibration control zone are synchronously collected through the electrical parameter monitoring unit to obtain the historical secondary voltage sequence and historical secondary current sequence of each vibration control zone in the historical monitoring period. Based on the historical secondary voltage sequence and historical secondary current sequence of each rapping control zone, the volt-ampere characteristic slope and flashover frequency at each sampling moment are calculated as indicators of the severity of back corona. According to the preset back corona intensity grading mapping rule, the indicators of the severity of back corona are converted into the back corona intensity corresponding to the current sampling moment, and the historical back corona intensity sequence of each rapping control zone in the historical monitoring period is generated. Based on the position coordinates of each of the aforementioned rapping control zones, the corresponding historical back corona intensity sequences are combined and arranged to generate a historical back corona intensity topology sequence for the dust collection electric field.

3. The method for suppressing back corona beat timing peak regulation according to claim 2, characterized in that, The steps for constructing the anti-corona intensity grading mapping rule include: The reference values ​​of the volt-ampere characteristic slope and flashover frequency of the electrostatic precipitator are obtained when it is operating normally under calibrated conditions. The reference values ​​of the volt-ampere characteristic slope and flashover frequency are rolled over according to a preset update cycle based on the real-time operating data of the electrostatic precipitator under non-back corona conditions. Based on the reference value of the current-voltage characteristic slope and the reference value of the flashover frequency, multiple current-voltage characteristic slope thresholds and multiple flashover frequency thresholds are set respectively, wherein the multiple current-voltage characteristic slope thresholds decrease sequentially along the direction of increasing back corona intensity, and the multiple flashover frequency thresholds increase sequentially along the direction of increasing back corona intensity. Based on the multiple volt-ampere characteristic slope thresholds and multiple flashover frequency thresholds, multiple continuous and non-overlapping back corona intensity level intervals are divided, and a unique back corona intensity level value is assigned to each back corona intensity level interval, forming a mapping relationship from the volt-ampere characteristic slope, the flashover frequency to the back corona intensity level value.

4. The method for controlling the timing of rapping to suppress back corona according to claim 1, characterized in that, Based on the historical back corona intensity topology sequence, the predicted back corona intensity sequence of each of the rapping control zones within a preset rapping control cycle is obtained, and the predicted back corona intensity topology sequence of the dust collection electric field is constructed, including: Using the historical monitoring cycle as the input time window length and the rapping control cycle as the output time window length, sample data is extracted from the historical operation monitoring records of the dust collection electric field. For each set of sample data, the anti-corona intensity topology sequence that satisfies the input time window length within a continuous time period is selected as the sample input sequence, and the anti-corona intensity topology sequence that is temporally adjacent to the sample input sequence and satisfies the output time window length is selected as the sample output sequence. The sample input sequence is paired with the sample output sequence to construct a sample dataset for training the prediction model; The Long Short-Term Memory network was trained to converge using the aforementioned sample dataset to generate an anti-corona intensity predictor. The historical back corona intensity topology sequence is input into the back corona intensity predictor, and the predicted back corona intensity topology sequence of the dust collection electric field during the rapping control cycle is output.

5. The method for controlling the timing of rapping to suppress back corona according to claim 1, characterized in that, Using the historical back corona intensity topology sequence, the predicted back corona intensity topology sequence, and the preset upper limit of back corona intensity as input, and with the rapping safety limit as constraint, and minimizing the impact of secondary dust and minimizing the amount of premature rapping as optimization objectives, the rapping sequence optimization is performed to solve for the optimal rapping timing scheme, including: Obtain a preset upper limit value for back corona intensity, and correct the upper limit value for back corona intensity downward according to a preset redundancy range to generate a limit value for back corona intensity. Based on the preset back corona intensity execution threshold, the historical back corona intensity topology sequence is filtered, and the vibration control partitions with historical back corona intensity greater than the back corona intensity execution threshold are selected as vibration partitions to be vibrated, thus determining the set of vibration partitions to be vibrated. Based on the predicted back corona intensity sequence corresponding to each of the partitions to be vibrated, calculate the latest allowed vibration triggering time under the premise that the predicted back corona intensity is always less than the back corona intensity limit value, and use it as the vibration cutoff time. Based on the predicted back corona intensity topology sequence and the set of partitions to be rapped, with the rapping trigger time of all partitions to be rapped not later than their respective rapping cutoff time as a hard constraint, and with the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, the rapping sequence optimization is performed to solve for the optimal rapping timing scheme.

6. The method for controlling the timing of rapping to suppress back corona according to claim 5, characterized in that, Based on the predicted back corona intensity topology sequence and the set of partitions to be rapped, with the rapping trigger time of all partitions to be rapped not later than their respective rapping cutoff time as a hard constraint, and with the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, the rapping sequence optimization is performed to solve for the optimal rapping timing scheme, including: The maximum number of simultaneous vibration triggers is obtained, and the number of vibration zones to be vibrated at the same time is less than or equal to the maximum number of simultaneous vibration triggers is used as a cooperative vibration constraint condition. Based on the set of partitions to be vibrated, under the premise of satisfying the cooperative vibration constraint, the vibration order of each partition to be vibrated is randomly arranged and the scheme is enumerated to generate several initial vibration timing schemes. Using the constraint that the vibration trigger time of all the partitions to be vibrated is no later than their respective vibration cutoff time as a hard constraint, the several initial vibration timing schemes are screened to determine several qualified vibration timing schemes. Based on the predicted back corona intensity topology sequence, with the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, the quality of the multiple qualified rapping timing schemes is evaluated, and the qualified rapping timing scheme corresponding to the scheme with the largest scheme quality coefficient is selected as the optimal rapping timing scheme.

7. The method for controlling the timing of rapping to suppress back corona according to claim 6, characterized in that, Based on the predicted back corona intensity topology sequence, with the optimization objectives of minimizing the impact of secondary dust and minimizing the amount of premature rapping, the quality of the multiple qualified rapping timing schemes is evaluated, including: For each qualified rapping timing scheme, based on the predicted back corona intensity topology sequence, the rapping trigger time of each partition to be rapped in the current scheme and the predicted back corona intensity value corresponding to the current time are extracted. Based on the rapping trigger time of each zone to be rapped, calculate the quantification value of the secondary dust impact of the current qualified rapping sequence scheme; Based on the predicted back corona intensity value of each zone to be vibrated and the back corona intensity limit value, calculate the advance vibration quantization value of the current qualified vibration timing scheme; After dimensionless processing of the quantified values ​​of the secondary dust impact and the quantified values ​​of the early rapping, the reciprocals of the two are weighted and summed to obtain the scheme quality coefficient of the current qualified rapping timing scheme.

8. The method for controlling the timing of rapping to suppress back corona according to claim 7, characterized in that, Based on the rapping trigger time of each zone to be rapped, calculate the quantification value of the secondary dust impact of the current qualified rapping sequence scheme, including: The vibration control cycle is divided into multiple consecutive discrete time slices; The number of partitions to be vibrated in the vibrating state within each discrete time slice is counted, a time slice vibrating number sequence is generated, and the maximum vibrating number of the time slice vibrating number sequence is determined. Calculate the mean number of rappings and the coefficient of variation of the number of rappings in the time-slice rapping sequence; The secondary dust impact quantification value of the current qualified rapping sequence scheme is determined based on the maximum number of rappings, the average number of rappings, and the coefficient of variation of the number of rappings. The secondary dust impact quantification value is positively correlated with the maximum number of rappings, the average number of rappings, and the coefficient of variation of the number of rappings.

9. The method for controlling the timing of rapping to suppress back corona according to claim 7, characterized in that, Based on the predicted back corona intensity value and the back corona intensity limit value for each zone to be rapped, the advance rapping quantization value of the current qualified rapping timing scheme is calculated, including: For each of the partitions to be rapped, the difference between the back corona intensity limit value and the predicted back corona intensity value of the current partition at the rapping trigger time is calculated as the advance rapping sub-quantity of the current partition. The advance rapping sub-quantities of all the sections to be rapped are summed, and the summation result is used as the advance rapping quantization value of the current qualified rapping timing scheme.

10. A rapping timing misalignment control system for suppressing back corona discharge, characterized in that, The method for implementing the rapping timing misalignment control method for suppressing back corona as described in any one of claims 1-9 includes: The partition configuration module is used to spatially divide the dust collection electric field of the electrostatic precipitator into several independent rapping control partitions, and to equip each partition with an independent rapping actuator and an electrical parameter monitoring unit. The back corona assessment module is used to collect the secondary voltage time-series data and secondary current time-series data of each of the rapping control zones in real time during operation through the electrical parameter monitoring unit, calculate the back corona severity index, determine the historical back corona intensity sequence of each of the rapping control zones in the historical monitoring period, and construct the historical back corona intensity topology sequence of the dust collection electric field. The back corona intensity assessment module is used to predict and obtain the predicted back corona intensity sequence of each of the rapping control zones within a preset rapping control cycle based on the historical back corona intensity topology sequence, and to construct the predicted back corona intensity topology sequence of the dust collection electric field. The rapping timing optimization module is used to take the historical back corona intensity topology sequence, the predicted back corona intensity topology sequence, and the preset back corona intensity upper limit as inputs, take the rapping safety limit as constraints, and take minimizing the impact of secondary dust and minimizing the amount of premature rapping as optimization objectives to perform rapping sequence optimization and solve the optimal rapping timing scheme. The vibration execution control module is used to trigger the vibration execution devices of each corresponding partition in chronological order according to the optimal vibration timing scheme.

Images