CN-122001271-A - Variable-frequency speed regulation energy-saving control method and system based on load dynamic identification

CN122001271ACN 122001271 ACN122001271 ACN 122001271ACN-122001271-A

Abstract

The invention relates to the technical field of variable frequency regulation, in particular to a variable frequency speed regulation energy-saving control method and a variable frequency speed regulation energy-saving control system based on load dynamic identification, wherein the method comprises the steps of constructing a multi-dimensional time sequence structure for data acquired in the current running period of a system, and identifying the dynamic trend of the current load based on the sequence; the method comprises the steps of establishing a candidate frequency set, outputting a power predicted value corresponding to the candidate frequency set through a power estimation model, calculating unit energy saving potential after being adjusted to the candidate frequency from the current frequency through an energy saving potential algorithm, inputting the candidate frequency corresponding to the unit energy saving potential which is larger than a threshold value into a feasible frequency set, establishing a comprehensive grading function on the basis of the feasible frequency set and the unit energy saving potential corresponding to the feasible frequency set, selecting an optimal target frequency through the comprehensive grading function, inputting the optimal target frequency and the current operating frequency into a feedback correction function, outputting the execution frequency of a frequency converter through the feedback correction function, and converting the execution frequency into a corresponding control instruction.

Inventors

CHEN TONGHUA
ZHOU QING
ZHANG ZHIXIONG
ZHAO YANZHI

Assignees

广东群策电机有限公司

Dates

Publication Date: 20260508
Application Date: 20260213

Claims (10)

1. The variable-frequency speed-regulating energy-saving control method based on load dynamic identification is characterized by comprising the following steps of: Constructing a multidimensional time sequence for the acquired data in the current operation period of the system, inputting the multidimensional time sequence into a combined structure of a convolution layer and an attention weighting module, and outputting a trend label through the combined structure; Establishing a candidate frequency set, outputting a power predicted value corresponding to the candidate frequency set through a power estimation model, calculating unit energy saving potential after being adjusted to the candidate frequency from the current frequency through an energy saving potential algorithm based on the power predicted value, and inputting the candidate frequency corresponding to the unit energy saving potential which is larger than a threshold value into a feasible frequency set, wherein the energy saving potential algorithm comprises an estimated value of average power in a latest period under the current frequency, the power predicted value of the candidate frequency and a trend weighting item, and the trend weighting item is used for automatically adjusting the sensitivity of an energy saving estimated value according to different trend states; On the basis of the feasible frequency set and the corresponding unit energy-saving potential thereof, a comprehensive scoring function is established according to double constraints of energy-saving efficiency and operation safety, the feasible frequency set is scored through the comprehensive scoring function to generate a scoring sequence, and the optimal target frequency is selected from a plurality of feasible frequencies; Inputting the optimal target frequency and the current operating frequency into a feedback correction function, outputting the executing frequency of the frequency converter through the feedback correction function, and converting the executing frequency into a corresponding control instruction, wherein the feedback correction function comprises a weighting result of the optimal target frequency and the current operating frequency and a frequency modulation executing weight factor; And acquiring the corresponding actual running power under the execution frequency, updating the energy-saving evaluation index, and providing dynamic feedback data for trend identification and energy-saving potential prediction of the next period.
2. The variable-frequency speed-regulation energy-saving control method based on load dynamic identification according to claim 1, wherein the converting the execution frequency into a corresponding control instruction specifically comprises: if the system adopts an analog quantity control mode, mapping the execution frequency into analog voltage or current through proportional conversion; If the digital communication mode is adopted, the execution frequency is converted into a set register value and written into the control address of the frequency converter.
3. The variable frequency speed and energy saving control method based on load dynamic identification according to claim 1, wherein the comprehensive scoring function comprises unit energy saving potential, a first penalty factor, a second penalty factor and a change rate of a current power mean value, wherein the first penalty factor is used for controlling the influence of frequency adjustment amplitude on scoring, and the second penalty factor is used for controlling risk caused by load response inertia.
4. The variable frequency speed regulation energy saving control method based on load dynamic identification according to claim 1, wherein the frequency modulation execution weight factor is used for mapping a system state into a frequency modulation speed, balances conflict between regulation response and stability, and comprises a comprehensive score corresponding to an optimal target frequency, an extremely small positive number for preventing zero removal, a power change rate of a current load, a trend label, an experience regulation coefficient and a weight correction coefficient.
5. The variable-frequency speed-regulation energy-saving control method based on load dynamic identification according to claim 1, wherein the energy-saving evaluation index represents a proportion of energy-saving effect brought by a current frequency modulation strategy, the energy-saving evaluation index is generated through an evaluation function, and the evaluation function comprises a reference power, an extremely small positive number for preventing denominator from being zero and a power evaluation index.
6. The variable frequency speed regulation energy saving control method based on load dynamic identification according to claim 5, wherein the power evaluation index is generated by combining an instantaneous power value with a frequency modulation execution weight factor, and the instantaneous power value is acquired by a three-phase power sensor installed at an output end of the motor.
7. The variable frequency speed regulation energy saving control method based on load dynamic identification according to claim 1, wherein the power estimation model consists of weighting terms of linear regression structure superposition trend factors.
8. The variable frequency speed control method based on dynamic load identification according to claim 1, wherein the trend label comprises load falling, load stabilizing and load rising.
9. The variable-frequency speed-regulating energy-saving control method based on load dynamic identification according to claim 1, wherein the collected data comprises the current running frequency of a frequency converter, motor current and real-time active power, the current running frequency is obtained through a communication register of the frequency converter, the motor current is collected by a current transformer arranged on a power supply side of the motor, the current is input to a controller after being conditioned by a signal, and the real-time active power can be obtained through internal parameters of the frequency converter or an independent power monitoring module.
10. Variable-frequency speed regulation energy-saving control system based on load dynamic identification, which is characterized by comprising: the trend identification module is used for constructing a multidimensional time sequence for the acquired data in the current running period of the system, inputting the multidimensional time sequence into a combined structure of the convolution layer and the attention weighting module, and outputting a trend label through the combined structure; The energy-saving potential evaluation module is used for establishing a candidate frequency set, outputting a power predicted value corresponding to the candidate frequency set through the power estimation model, calculating the unit energy-saving potential after being adjusted from the current frequency to the candidate frequency through an energy-saving potential algorithm based on the power predicted value, and inputting the candidate frequency corresponding to the unit energy-saving potential which is larger than a threshold value into a feasible frequency set, wherein the energy-saving potential algorithm comprises an estimated value of average power in a latest period under the current frequency, the power predicted value of the candidate frequency and a trend weighting item, and the trend weighting item is used for automatically adjusting the sensitivity of an energy-saving estimated value according to different trend states; The target frequency generation module is used for establishing a comprehensive scoring function according to double constraints of energy saving efficiency and operation safety on the basis of the feasible frequency set and the corresponding unit energy saving potential thereof, scoring the feasible frequency set through the comprehensive scoring function to generate a scoring sequence, and selecting an optimal target frequency from a plurality of feasible frequencies; The frequency correction model is used for inputting the optimal target frequency and the current running frequency into a feedback correction function, outputting the execution frequency of the frequency converter through the feedback correction function, and converting the execution frequency into a corresponding control instruction, wherein the feedback correction function comprises a weighting result of the optimal target frequency and the current running frequency and a frequency modulation execution weighting factor; And the index updating module is used for updating the energy-saving evaluation index based on the execution frequency and providing dynamic feedback data for trend identification and energy-saving potential prediction of the next period.

Description

Variable-frequency speed regulation energy-saving control method and system based on load dynamic identification Technical Field The invention relates to the technical field of variable frequency regulation, in particular to a variable frequency speed regulation energy-saving control method and system based on load dynamic identification. Background In energy-saving control of motor loads, variable frequency speed regulation has become a mainstream means, but engineering sites still commonly adopt regulation modes based on threshold triggering or fixed rules, more depend on instantaneous quantity or experience curves, and are difficult to describe dynamic characteristics such as rising, falling and fluctuation of loads in a short time. The common problems of the method include that firstly, only the current state is seen, the upcoming change direction and change speed are not recognized, so that the frequency modulation time is inaccurate, secondly, the frequency modulation action is regarded as zero cost, the cost of frequency change and equipment inertia are not brought into calculation, frequent and small step sizes are easy to occur, and constraint adjustment is easy to bring about system fluctuation, mechanical impact and energy loss, thirdly, energy efficiency evaluation is independent of a control chain and is often based on rough statistical average or post summary, quantization indexes which are useful for the next period and can be directly utilized are difficult to form, fourthly, the execution level adopts a direct writing target frequency mode, the influence of acceleration and deceleration paths on motor current, pipe network hydraulic or pneumatic system stability under different load scenes is ignored, and fifthly, the data sources are scattered, the time organization is loose, the history window is not used in a structured mode, and the reusable running state expression cannot be formed. As a result, the system is easy to have the problems of unstable energy-saving effect, unmatched adjusting action and physical constraint, unexplained strategy, difficult convergence in long-term operation and the like in complex working conditions, and the further improvement of frequency conversion and energy saving is restricted. For the pain point, a system method is needed, which takes a real-time state as a basis, takes trend identification as an entrance, takes common measurement of energy saving benefits and physical cost as a core and takes an executable path as an exit, and the identification, evaluation, decision and execution are sequentially connected in an engineering manner, so that stable and reusable energy saving control is realized on the premise of not changing the existing hardware basis. Disclosure of Invention The invention aims to provide a variable-frequency speed regulation energy-saving control method and system based on load dynamic identification, so as to solve the problems in the background technology. In order to achieve the above purpose, the invention adopts the following technical scheme: A variable-frequency speed-regulating energy-saving control method based on load dynamic identification comprises the following steps: Constructing a multidimensional time sequence for the acquired data in the current operation period of the system, inputting the multidimensional time sequence into a combined structure of a convolution layer and an attention weighting module, and outputting a trend label through the combined structure; Establishing a candidate frequency set, outputting a power predicted value corresponding to the candidate frequency set through a power estimation model, calculating unit energy saving potential after being adjusted to the candidate frequency from the current frequency through an energy saving potential algorithm based on the power predicted value, and inputting the candidate frequency corresponding to the unit energy saving potential which is larger than a threshold value into a feasible frequency set, wherein the energy saving potential algorithm comprises an estimated value of average power in a latest period under the current frequency, the power predicted value of the candidate frequency and a trend weighting item, and the trend weighting item is used for automatically adjusting the sensitivity of an energy saving estimated value according to different trend states; On the basis of the feasible frequency set and the corresponding unit energy-saving potential thereof, a comprehensive scoring function is established according to double constraints of energy-saving efficiency and operation safety, the feasible frequency set is scored through the comprehensive scoring function to generate a scoring sequence, and the optimal target frequency is selected from a plurality of feasible frequencies; Inputting the optimal target frequency and the current operating frequency into a feedback correction function, outputting the executing frequency of the frequency c