An intelligent control system for frequency converter operation

Abstract

The application discloses a kind of frequency converter operation intelligent control systems, belong to frequency converter control technical field, including platform end and control end;Platform end includes algorithm library, load analysis module;Control end includes acquisition module, control analysis module;Algorithm library is used to store the various reserve algorithms of control analysis to frequency converter;Load analysis module is used to reserve analysis in real time based on algorithm library to the frequency converter of user's load equipment, obtains the control strategy of frequency converter, sends control strategy to control end;Acquisition module is used to carry out real-time data acquisition, obtains corresponding working condition acquisition data, carries out feature recognition to working condition acquisition data, obtains corresponding working condition feature set;Control analysis module is used to carry out control analysis, identifies working condition feature set in real time, according to preset control strategy, working condition feature set is analyzed, obtains corresponding analysis result, controls frequency converter according to analysis result.

Description

Technical Field

[0001] This invention belongs to the field of frequency converter control technology, specifically an intelligent control system for frequency converter operation. Background Technology

[0002] In industrial applications, the frequency converter control system is the core device for regulating motor operation. However, in actual operation, the types of loads driven by the frequency converter are extremely diverse, far beyond what a single motor can cover, and different loads have significantly different performance requirements for the frequency converter.

[0003] Taking fans as an example, their operating characteristics dictate that their speed must be strictly controlled within a specific frequency range to maintain optimal efficiency. If the frequency range set by the frequency converter does not match the fan's design parameters, the fan will struggle to operate stably at its optimal operating point. This not only leads to a significant reduction in the fan's efficiency and a substantial increase in energy consumption, but also accelerates the wear and tear on the motor and fan due to prolonged operation in a suboptimal state. Similar situations are prevalent in other load types, such as water pumps. Their flow rate and head are closely related to their speed. If the frequency converter cannot accurately control the pump speed to operate within the appropriate frequency range, the pump will fail to provide the required flow rate and head, affecting the normal operation of the entire water system. In heating systems, excessively low pump speeds can lead to insufficient heating; while in industrial circulating water systems, excessively high pump speeds can result in energy waste and increased pipeline pressure, posing a threat to system safety.

[0004] Based on this, in order to achieve intelligent control of inverter operation, the present invention provides an intelligent control system for inverter operation. Summary of the Invention

[0005] To address the problems of the above solutions, this invention provides an intelligent control system for inverter operation.

[0006] The objective of this invention can be achieved through the following technical solutions: An intelligent control system for inverter operation includes a platform and a control terminal; Furthermore, there is a communication connection between the platform and the control terminal.

[0007] The platform includes an algorithm library and a load analysis module; The algorithm library is used to store various reserve algorithms for controlling and analyzing frequency converters.

[0008] Furthermore, the reserve algorithm includes a fusion algorithm.

[0009] The load analysis module is used to perform real-time reserve analysis on the inverters of the user's load equipment based on the algorithm library, obtain the control strategy of the inverter, and send the control strategy to the control terminal.

[0010] Furthermore, based on the algorithm library, real-time backup analysis is performed on the inverters of the user's load equipment, including: Obtain the user's historical frequency converter data, identify various operating condition characteristics based on the historical frequency converter data, perform feature identification on the historical frequency converter data based on each operating condition characteristic, and obtain the single operating condition range corresponding to each operating condition characteristic. Dynamically classify each individual working condition range to obtain several working condition classifications; generate control strategies based on each working condition classification.

[0011] Furthermore, dynamic classification is performed based on the range of each individual operating condition, including: A digital twin of the load device is established, and the digital twin is used to simulate each individual working condition range in real time to obtain the simulation results corresponding to the corresponding working condition feature set. The simulation results of the same working condition feature set under each reserve algorithm are identified, and each simulation result is filtered to obtain the simulation representative result corresponding to each working condition feature set. The simulation representative result is labeled with the corresponding reserve algorithm label. Based on whether the simulation results are the same, the feature sets of each working condition are merged to obtain several initial classifications; The results of the simulated representative results of each initial category are used to conduct a merge evaluation to obtain the merge evaluation results between the corresponding initial categories. The merge evaluation results include merge qualified and merge unqualified. The merge assessment is combined into a single initial classification that meets the merge criteria, resulting in a new initial classification, and a simulated representative result for the new initial classification is determined. This process continues until no more qualified initial categories are merged, at which point each initial category is marked as a working condition category.

[0012] Furthermore, the initial classifications are evaluated by merging based on the simulation results, including: Establish a merger assessment model, the expression of which is: ; In the formula: (q, p) are the input data, q and p represent the two simulation results to be merged and evaluated, q↔p means that the two simulation results meet the merging requirements; the output data is the merged evaluation value HP(q, p), and the merged evaluation value is 1 or 0; The combined evaluation model is used to analyze the results of the two simulations to obtain the combined evaluation values; the combined evaluation results between the initial categories are then determined based on the combined evaluation values.

[0013] Furthermore, the control strategy is as follows: Real-time identification of the inverter's operating condition feature set, matching the corresponding operating condition classification based on the operating condition feature set, and identifying the reserve algorithm corresponding to the operating condition classification; Priority analysis is performed on each reserve algorithm, and the reserve algorithm with the highest priority is selected as the target algorithm. The target algorithm is then analyzed to obtain the corresponding analysis results.

[0014] Furthermore, control strategies are generated based on various operating conditions, including: The platform establishes a working condition prediction model, performs real-time working condition prediction through the working condition prediction model, and obtains the working condition prediction feature set at the corresponding time. Based on the working condition prediction feature set, it matches the corresponding working condition classification and identifies the reserve algorithm corresponding to the working condition classification. The various reserve algorithms and working condition prediction feature sets are simulated using a digital twin model to obtain the simulation prediction results corresponding to each reserve algorithm. The simulation prediction results corresponding to each reserve algorithm are compared, and the reserve algorithm with the best performance is selected as the target algorithm for the working condition prediction feature. The system identifies the operating condition feature set of the frequency converter in real time, matches the operating condition feature set with the operating condition prediction feature set, analyzes the matched operating condition prediction feature set according to the target algorithm, and obtains the corresponding analysis results.

[0015] The control terminal includes a data acquisition module and a control analysis module; The acquisition module is used to acquire data in real time, obtain corresponding operating condition data, perform feature recognition on the operating condition data, and obtain corresponding operating condition feature sets.

[0016] The control analysis module is used to perform control analysis, identify the operating condition feature set in real time, analyze the operating condition feature set according to the preset control strategy, obtain the corresponding analysis results, and control the frequency converter based on the analysis results.

[0017] Compared with the prior art, the beneficial effects of the present invention are: The intelligent control system for inverter operation of this invention can deeply analyze the unique operating characteristics of different load types. Taking a fan as an example, the system can accurately identify the specific frequency range within which its speed needs to be strictly controlled, ensuring a perfect match between the frequency set by the inverter and the fan. This allows the fan to always operate stably at its optimal operating point, significantly improving the fan's efficiency. For a water pump, the system can precisely control the pump speed based on the close relationship between its flow rate, head, and speed, allowing it to operate within a suitable frequency range. This effectively provides the required flow rate and head, ensuring the efficient operation of the entire water system and avoiding inefficiencies caused by frequency mismatch. Optimized energy utilization: Through intelligent control of the inverter operation, the system can dynamically adjust the output frequency and power according to the actual needs of the load. This avoids unnecessary energy consumption during the operation of loads such as fans and water pumps. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a block diagram illustrating the principle of the present invention. Detailed Implementation

[0020] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0021] like Figure 1 As shown, an intelligent control system for inverter operation includes a platform and a control terminal. The platform and the control terminal are connected via a communication connection.

[0022] The platform includes an algorithm library and a load analysis module; The algorithm library is used to store various reserve algorithms for controlling and analyzing frequency converters.

[0023] The platform can configure various reserve algorithms, such as fuzzy control, neural network control, and expert system control. It can also be a fusion algorithm of various algorithms, adjusting the weights of the corresponding algorithms. For example, fuzzy control and neural network control algorithms can be run simultaneously, and the weights of the two algorithms can be adjusted in real time according to changes in working conditions.

[0024] The load analysis module is used to perform real-time reserve analysis on the inverters of the user's load equipment based on the algorithm library, obtain the control strategy of the inverter, and send the control strategy to the control terminal.

[0025] In one embodiment, reserve analysis can be performed based on digital twin technology. Dynamic simulation is conducted according to the actual situation of the frequency converter to determine the reserve algorithm to be applied under various operating conditions. If it is a fusion algorithm, it has corresponding algorithms and related data such as weights. The data are integrated to form a control strategy, and then the corresponding reserve algorithm can be applied for analysis and processing according to the actual operating conditions.

[0026] In one embodiment, real-time reserve analysis of the inverters of the user's load equipment is performed based on an algorithm library, including: The system acquires historical inverter data from users, including historical motor operating parameters (speed, current, voltage, temperature, etc.), operating characteristics (statistics such as mean, variance, and peak values ​​of speed, current, and voltage, as well as external factors such as load change rate and ambient temperature), control strategies (including the process of selecting reserve algorithms), control effects, and effect evaluation results. This data can be transmitted in real time from the control terminal or periodically.

[0027] Based on historical frequency converter data, various operating condition characteristics are identified, such as the mean, variance, and peak values ​​of speed, current, and voltage. Alternatively, the platform can manually set these characteristics initially. Based on each operating condition characteristic, the historical frequency converter data is characterized to obtain the data range of each operating condition characteristic and mark it as a single operating condition range. Dynamic classification is performed based on the range of each individual working condition to obtain several working condition classifications; that is, when the working condition is in the corresponding working condition classification, any of the reserve algorithms corresponding to the working condition classification can be applied. Control strategies are generated based on the classification of various operating conditions.

[0028] In one embodiment, dynamic classification is performed based on individual operating condition ranges. This classification is based on existing methods and adjusted according to changes in load equipment and frequency converters. For example, a fresh air fan impeller with a smooth surface and regular shape exhibits good aerodynamic performance and maintains high efficiency within a specific frequency range. However, with prolonged use, the impeller may experience surface roughness and deformation due to wear and corrosion, leading to a decline in aerodynamic performance. A fan that initially operated efficiently in a certain frequency range may need to be adjusted to other frequency ranges to achieve relatively good efficiency due to impeller issues. This renders the previously defined optimal operating condition classification inapplicable, necessitating adjustments to the operating condition classification.

[0029] In one embodiment, dynamic classification is performed based on the range of each individual operating condition, including: A digital twin of the load equipment is established based on digital twin technology, including the corresponding frequency converter; the digital twin is used to simulate each individual working condition range in real time, and the simulation results of the simulation under the working condition feature set formed by the digital twin under each working condition feature item and the reserve algorithm are obtained. The working condition feature set is the set of working condition features corresponding to each working condition feature item. Identify the simulation results of the same working condition feature set under various reserve algorithms, filter each simulation result, including the best-performing simulation result, and remove the others to obtain the representative simulation result corresponding to each working condition feature set; alternatively, prioritize each simulation result based on the negotiation priority algorithm and select the highest priority as the representative simulation result; and label the representative simulation result with the corresponding reserve algorithm label. Based on whether the simulation representative results are the same, the feature sets of each working condition are merged to obtain several initial classifications. That is, the working condition feature sets with the same simulation representative results are merged. For example, if the working condition feature sets with the same simulation representative results are (1, 2, 3) and (1, 5, 6), then the initial classification is (1, 2 and 5, 3 and 6), which can be regarded as the union of the corresponding working condition feature items. The working condition feature sets without the same results are independently used as the initial classifications. Based on the results of each simulation representative, each initial category is merged and evaluated to obtain the merged evaluation result, which includes merged qualified and merged unqualified. The merged assessment is combined into a single initial classification that meets the merge criteria, resulting in a new initial classification. The simulated representative result of this new initial classification can be an intermediate result between the two initial classifications or a set result, which includes the simulated representative results corresponding to the unmerged classifications. In this case, the merge is considered qualified only if it meets the merge requirements with each of the simulated representative results. This process continues until no more qualified initial categories are merged, at which point each initial category is marked as a working condition category.

[0030] In one embodiment, if the historical frequency converter data has complete control effects, effect evaluation results, etc., and the data is sufficient, it is possible to directly identify the data without using a digital twin for simulation.

[0031] In one embodiment, the simulation results are filtered, preset result evaluation items are set, the simulation results are evaluated according to each result evaluation item to obtain the corresponding simulation score, and the results are filtered according to the size of the simulation score.

[0032] For example, each individual item score can be evaluated according to the evaluation indicators of each outcome evaluation item, and then a simulated score can be calculated based on the weight coefficient of each outcome evaluation item. Alternatively, other scoring methods can be used for evaluation.

[0033] In one embodiment, the initial classifications are evaluated by merging based on the simulation results. This evaluation can be based on similarity, i.e., if the similarity is greater than a preset value; or it can be based on intelligent models built using machine learning, deep learning algorithms, etc., for intelligent evaluation.

[0034] In one embodiment, the initial classifications are evaluated by merging based on the simulation results, including: The platform uses historical data to label whether various simulation results meet the merging requirements, forming a training set. Based on this training set, a merging evaluation model is built. The expression for the merging evaluation model is: ; In the formula: (q, p) are the input data, q and p represent the two simulation results to be merged and evaluated, q↔p means that the two simulation results meet the merging requirements; the output data is the merged evaluation value HP(q, p), and the merged evaluation value is 1 or 0; The two simulation results are analyzed by merging the evaluation model to obtain the corresponding merged evaluation value. If the initial classification has multiple simulation results, each simulation result is analyzed with each simulation result of the other initial classification. If any of the merged evaluation values ​​is 0, the two simulation results cannot be merged. Determine whether the initial categories are eligible for merging based on the respective merge evaluation values ​​corresponding to the initial categories.

[0035] In one embodiment, a control strategy is generated based on each operating condition classification. The generation can be carried out in an optional manner, such as real-time identification of the operating condition category to which the operating condition feature set belongs, and then selecting one of the reserve algorithms corresponding to the operating condition category for application.

[0036] In one embodiment, the difference between this embodiment and the previous embodiment is that the optional reserve algorithms are not evaluated in an optional manner, and the one with the highest priority is selected for application. For example, the priority is ranked based on single analysis efficiency, analysis effect, analysis accuracy, etc., and multiple parameters can also be analyzed comprehensively.

[0037] In one embodiment, a control strategy is generated based on various operating conditions, including: The platform establishes a working condition prediction model to predict subsequent working condition feature sets. This prediction utilizes existing prediction technologies and can also be combined with digital twins, such as building a working condition prediction model based on deep learning algorithms. Real-time working condition prediction is performed using this model to obtain the working condition prediction feature set for the corresponding time period. This means outputting all possible features, or only outputting the most probable feature set to reduce the analysis workload, but the prediction accuracy must meet preset requirements. Based on the working condition prediction feature set, the platform matches the corresponding working condition classification and identifies the reserve algorithms corresponding to each classification. The simulation of each reserve algorithm and the working condition prediction feature set is carried out using a digital twin model to obtain the simulation prediction results corresponding to each reserve algorithm. That is, simulation is carried out according to the working condition prediction feature set and the reserve algorithm to obtain the corresponding analysis results, and control is carried out according to the analysis results to obtain the simulation prediction results. The simulation prediction results corresponding to each reserve algorithm are compared, and the reserve algorithm with the best effect is selected as the target algorithm for the corresponding working condition prediction feature set. The system identifies the operating condition feature set of the frequency converter in real time, matches the operating condition feature set with the operating condition prediction feature set, analyzes the matched operating condition prediction feature set according to the target algorithm, obtains the corresponding analysis results, and integrates them into the control strategy.

[0038] The control terminal includes a data acquisition module and a control analysis module; The acquisition module is used to acquire data in real time, obtain corresponding operating condition data, perform feature recognition on the operating condition data, and obtain corresponding operating condition feature sets.

[0039] For example, low-cost, high-precision sensors can be used to monitor parameters such as motor speed, current, voltage, and temperature in real time, while other smart sensors can collect external operating condition data, such as load changes, ambient temperature, and power grid voltage fluctuations.

[0040] The collected raw data is filtered to eliminate noise and interference and improve data quality; the data is normalized to convert data of different dimensions to the same scale for easier processing by subsequent algorithms; key feature parameters, such as the rate of change of speed and the range of current fluctuations, are extracted, which can reflect the real-time operating status and changes in operating conditions of the motor; based on the preprocessed data, feature parameters that can characterize the current operating conditions are extracted; these feature parameters may include statistical quantities such as the mean, variance, and peak value of speed, current, and voltage, as well as external factors such as the rate of change of load and ambient temperature.

[0041] The control analysis module is used to perform control analysis, identify the operating condition feature set in real time, analyze the operating condition feature set according to the preset control strategy, obtain the corresponding analysis results, and control the frequency converter based on the analysis results.

[0042] The above formulas are all numerical calculations after removing dimensions. The formulas are obtained by software simulation based on a large amount of data and are closest to the real situation. The preset parameters and preset thresholds in the formulas are set by those skilled in the art according to the actual situation or obtained by simulation based on a large amount of data.

[0043] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. An intelligent control system for inverter operation, characterized in that, Includes the platform side and the control side; The platform includes an algorithm library and a load analysis module; the control terminal includes a data acquisition module and a control analysis module. The algorithm library is used to store various reserve algorithms for controlling and analyzing frequency converters; The load analysis module is used to perform real-time reserve analysis on the inverters of the user's load equipment based on the algorithm library, obtain the control strategy of the inverter, and send the control strategy to the control terminal. The acquisition module is used to acquire data in real time, obtain corresponding working condition data, and perform feature recognition on the working condition data to obtain corresponding working condition feature sets. The control analysis module is used to perform control analysis, identify the operating condition feature set in real time, analyze the operating condition feature set according to the preset control strategy, obtain the corresponding analysis results, and control the frequency converter based on the analysis results.

2. The intelligent control system for inverter operation according to claim 1, characterized in that, The platform and the control terminal are connected via a communication connection.

3. The intelligent control system for inverter operation according to claim 1, characterized in that, The reserve algorithm includes a fusion algorithm.

4. The intelligent control system for inverter operation according to claim 1, characterized in that, Based on an algorithm library, real-time backup analysis is performed on the inverters of the user's load equipment, including: Obtain the user's historical frequency converter data, identify various operating condition characteristics based on the historical frequency converter data, perform feature identification on the historical frequency converter data based on each operating condition characteristic, and obtain the single operating condition range corresponding to each operating condition characteristic. Dynamically classify each individual working condition range to obtain several working condition classifications; generate control strategies based on each working condition classification.

5. The intelligent control system for inverter operation according to claim 4, characterized in that, Dynamic classification is performed based on the range of each individual operating condition, which is based on existing methods and includes: A digital twin of the load device is established, and the digital twin is used to simulate each individual working condition range in real time to obtain the simulation results corresponding to the corresponding working condition feature set. The simulation results of the same working condition feature set under each reserve algorithm are identified, and each simulation result is filtered to obtain the simulation representative result corresponding to each working condition feature set. The simulation representative result is labeled with the corresponding reserve algorithm label. Based on whether the simulation results are the same, the feature sets of each working condition are merged to obtain several initial classifications; The results of the simulated representative results of each initial category are used to conduct a merge evaluation to obtain the merge evaluation results between the corresponding initial categories. The merge evaluation results include merge qualified and merge unqualified. The merge assessment is combined into a single initial classification that meets the merge criteria, resulting in a new initial classification, and a simulated representative result for the new initial classification is determined. This process continues until no more qualified initial categories are merged, at which point each initial category is marked as a working condition category.

6. The intelligent control system for inverter operation according to claim 5, characterized in that, The initial classifications are evaluated by merging based on the simulation results, including: Establish a merger assessment model, the expression of which is: ； In the formula: (q, p) are the input data, q and p represent the two simulation results to be merged and evaluated, q↔p means that the two simulation results meet the merging requirements; the output data is the merged evaluation value HP(q, p), and the merged evaluation value is 1 or 0; By analyzing the two simulation results using a combined evaluation model, the corresponding combined evaluation values ​​are obtained; based on the combined evaluation values, the combined evaluation results between the corresponding initial categories are determined.

7. The intelligent control system for inverter operation according to claim 4, characterized in that, The control strategy is as follows: Real-time identification of the inverter's operating condition feature set, matching the corresponding operating condition classification based on the operating condition feature set, and identifying the reserve algorithm corresponding to the operating condition classification; Priority analysis is performed on each reserve algorithm, and the reserve algorithm with the highest priority is selected as the target algorithm. The target algorithm is then analyzed to obtain the corresponding analysis results.

8. The intelligent control system for inverter operation according to claim 4, characterized in that, Control strategies are generated based on various operating conditions, including: The platform establishes a working condition prediction model, performs real-time working condition prediction through the working condition prediction model, and obtains the working condition prediction feature set at the corresponding time. Based on the working condition prediction feature set, it matches the corresponding working condition classification and identifies the reserve algorithm corresponding to the working condition classification. The various reserve algorithms and working condition prediction feature sets are simulated using a digital twin model to obtain the simulation prediction results corresponding to each reserve algorithm. The simulation prediction results corresponding to each reserve algorithm are compared, and the reserve algorithm with the best performance is selected as the target algorithm for the working condition prediction feature. The system identifies the operating condition feature set of the frequency converter in real time, matches the operating condition feature set with the operating condition prediction feature set, analyzes the matched operating condition prediction feature set according to the target algorithm, and obtains the corresponding analysis results.

Images