Motor energy-saving system and method based on load identification algorithm

Abstract

This invention relates to the field of motor control and energy-saving technology, specifically to a motor energy-saving system based on a load identification algorithm. The system includes: a load state sensing module connected to the motor power supply circuit; a dynamic load feature construction module for extracting load feature data based on the motor's operating electrical parameters, the load feature data containing multi-dimensional feature information characterizing the motor load's state over time; a load identification decision module for identifying the current motor load state and outputting a load state label; and a motor drive control module for determining target control parameters based on the load state label and adjusting the motor's input voltage or input frequency. This invention employs a modular closed-loop design, and the load identification model uses an offline training and online inference architecture, resulting in low computational overhead. It is suitable for deployment in embedded platforms and existing motor control systems, exhibiting good versatility and engineering applicability.

Description

Technical Field

[0001] This invention relates to the field of motor control and energy-saving technology, specifically to a motor energy-saving system and method based on a load identification algorithm. Background Technology

[0002] In existing motor energy-saving control schemes, some technologies employ fixed parameter control or constant voltage / constant frequency control methods. These control strategies are suitable for operating scenarios with relatively stable loads. However, in practical applications, motor loads often exhibit periodic or random fluctuations. When the motor operates in a non-rated load range for extended periods, the matching degree between the fixed parameter control strategy and the current operating conditions needs further improvement.

[0003] Regarding the acquisition of load status, there are two main methods in the existing technology. One is to indirectly measure the load by using external mechanical sensors. This method can obtain certain load information, but the system hardware cost is high, and the sensor signal is easily affected by the installation method and environmental factors.

[0004] Another method is to infer the load based on the motor's operating electrical parameters. This method does not require additional sensors, but the utilization of electrical parameters is mostly limited to the threshold judgment of a single electrical quantity, and the dynamic process information of the load changing over time is not fully utilized.

[0005] In summary, how to make fuller use of the dynamic changes in the motor's own operating electrical parameters to accurately identify the load state without increasing hardware costs, and dynamically match the motor input voltage based on the identification results, so as to improve the overall operating efficiency of the motor under variable load conditions, is a research direction worthy of attention in the field of motor energy-saving control. Therefore, we propose a motor energy-saving system and method based on load identification algorithm. Summary of the Invention

[0006] According to embodiments of the present invention, a motor energy-saving system and method based on a load identification algorithm are provided to address the problems existing in the present invention.

[0007] This motor energy-saving system based on a load identification algorithm includes:

[0008] Load status sensing module, dynamic load characteristic construction module, load identification and decision module, and motor drive control module;

[0009] A load status sensing module is connected to the motor power supply circuit to collect motor operating electrical parameters in real time.

[0010] A dynamic load feature construction module is connected to the load state sensing module and is used to extract load feature data based on the motor operating electrical parameters. The load feature data includes multi-dimensional feature information characterizing the motor load change state over time.

[0011] A load identification decision module, which is connected to the dynamic load feature construction module, is used to input the load feature data into a pre-trained load identification model to identify the current motor load state and output a load state label.

[0012] A motor drive control module, which is connected to the load identification and decision module, is used to determine the target control parameters based on the load status label and adjust the input voltage or input frequency of the motor.

[0013] Preferably, the dynamic load characteristic construction module includes:

[0014] The time-series sampling submodule is used to continuously sample the motor operating electrical parameters within a preset detection period to obtain a time series of motor operating parameters;

[0015] The feature extraction submodule is used to extract features from the time series of motor operating parameters to generate the load feature data.

[0016] Preferably, the load identification decision module has a built-in load identification model trained on working condition samples. The load identification model is used to receive the load feature data and output the corresponding load status label. The load status label includes light load status, medium load status, heavy load status and overload status.

[0017] Preferably, the motor drive control module includes:

[0018] The parameter parsing submodule is used to query a preset parameter mapping table based on the load status label in order to determine the target control parameter;

[0019] The drive execution submodule is used to generate drive control signals based on the target control parameters and to dynamically adjust the input voltage of the motor.

[0020] Preferably, the drive execution submodule uses voltage regulation to adjust the motor input voltage, and the target regulation parameter is a voltage regulation command, which includes voltage amplitude adjustment amount and voltage regulation rate.

[0021] Preferably, the motor operating electrical parameters include the motor's three-phase current and three-phase voltage. The dynamic load feature construction module constructs the load feature data by extracting and combining features from the three-phase current and three-phase voltage.

[0022] Preferably, the load identification decision module is further configured to compare the identified load status label with the historical load status label, and when the load status changes across levels, output a status switching indicator signal, and the motor drive control module responds to the status switching indicator signal to perform fast response control.

[0023] Preferably, the load identification model is a classification model based on statistical learning or neural networks, which is obtained through offline training and keeps the model parameters unchanged during system operation, outputting the load status label with fixed decision logic.

[0024] An implementation of a motor energy-saving system based on a load identification algorithm according to any one of claims 1 to 8 includes:

[0025] S1: The load status sensing module collects the motor's operating electrical parameters in real time, and generates standardized current timing data and voltage timing data after internal preprocessing;

[0026] S2: Receive the current timing data and voltage timing data through the dynamic load feature construction module, extract the dynamic change features of the load through a sliding time window, and generate load feature data;

[0027] S3: The load feature data is input into the pre-trained load recognition model through the load recognition decision module. The model then infers and identifies the current motor load state and outputs the corresponding load state label.

[0028] S4: Receive the load status tag through the motor drive control module, match the corresponding target control parameters according to the load status tag, and dynamically adjust the input voltage of the motor according to the target control parameters to achieve energy-saving operation of the motor.

[0029] Preferably, the offline training process of the load identification model in step S3 includes: collecting samples of the motor's operating electrical parameters under various known load conditions, extracting load feature data from the operating electrical parameter samples and labeling them with corresponding load state labels, inputting the labeled load feature data as a training set into the classification model to be trained, and iteratively optimizing the classification model to learn the mapping relationship between the motor's operating electrical parameters and the load state.

[0030] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0031] This invention employs a modular closed-loop design, with the load identification model using an offline training and online inference architecture. This results in low computational overhead, making it suitable for deployment in embedded platforms and existing motor control systems. It also demonstrates good versatility and engineering applicability.

[0032] The dynamic load feature construction module performs time-series sampling and multi-dimensional feature extraction of the motor's three-phase current and voltage to generate load feature data characterizing the load's state over time. Compared to methods relying on single electrical quantity thresholds or indirect measurements from external sensors, this improves the accuracy of motor load state perception without increasing hardware costs. Simultaneously, the load identification decision module compares the current identification result with historical tags. When a load level transition occurs, it outputs a state switching indicator signal and triggers the drive execution submodule response, improving the timeliness of the linkage between load identification and control regulation.

[0033] The system queries a preset mapping table based on the load status label to determine the target control parameters, including voltage amplitude adjustment and adjustment rate, and dynamically adjusts the motor input voltage. Under light or medium load conditions, the voltage is reduced to minimize iron and copper losses; under heavy or overload conditions, the rated voltage is restored to ensure output. This achieves adaptive voltage regulation of the motor across the entire operating range of light, medium, heavy, and overload conditions, balancing operational safety and energy efficiency.

[0034] It should be understood that the description in the Summary of the Invention is not intended to limit the key or essential features of the embodiments of the present invention, nor is it intended to restrict the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0035] The above and other features, advantages, and aspects of the various embodiments of the present invention will become more apparent from the accompanying drawings and the following detailed description. In the drawings, the same or similar reference numerals denote the same or similar elements, wherein:

[0036] Figure 1 A schematic diagram of the system flow of a motor energy-saving system and method based on a load identification algorithm according to the present invention is shown.

[0037] Figure 2 A schematic flowchart of a motor energy-saving system and method based on a load identification algorithm according to the present invention is shown. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0040] like Figure 1 and Figure 2 As shown, the motor energy-saving system based on load identification algorithm includes a load status sensing module, a dynamic load feature construction module, a load identification decision module, and a motor drive control module.

[0041] The load status sensing module is connected to the motor power supply circuit, and its output is connected to the input of the dynamic load characteristic construction module. The output of the dynamic load characteristic construction module is connected to the input of the load identification and decision module, and the output of the load identification and decision module is connected to the control terminal of the motor drive control module. The power output terminal of the motor drive control module is electrically connected to the motor power supply circuit, forming a closed-loop control link.

[0042] The load status sensing module continuously collects operating electrical parameters such as three-phase current and three-phase voltage of the motor during motor operation. The collected analog signals are filtered and amplified by the internal signal conditioning circuit, and then converted into digital current timing data and voltage timing data by the analog-to-digital conversion unit.

[0043] After receiving the aforementioned time-series data, the dynamic load feature construction module continuously samples the current and voltage time-series data within a preset detection period to obtain a time series of motor operating parameters. Then, it uses a feature extraction algorithm to perform time-domain feature calculations on this time series to generate load feature data. This load feature data contains multi-dimensional feature information characterizing the motor load's state over time.

[0044] The load identification decision module has a pre-built load identification model trained offline, which is constructed based on classification algorithms such as Gaussian mixture models or support vector machines. The load feature data output by the dynamic load feature construction module is fed into the load identification model as input. After inference calculation, the model outputs the corresponding load state label. The load state label is divided into multiple levels according to the actual operating conditions of the motor, such as light load, medium load, heavy load, and overload.

[0045] The load status tag output by the load identification and decision module is sent to the motor drive control module. The motor drive control module internally includes a parameter parsing submodule and a drive execution submodule. The parameter parsing submodule pre-stores a mapping table between load status and control parameters. It queries this mapping table based on the received load status tag to determine the target control parameter matching the load status and outputs it as a voltage regulation command.

[0046] The voltage regulation command includes the voltage amplitude adjustment amount and the voltage regulation rate. The drive execution submodule generates a corresponding drive control signal based on the voltage regulation command, and dynamically adjusts the motor input voltage amplitude without changing the motor power supply frequency, thereby achieving energy-saving operation while ensuring normal motor operation.

[0047] During system operation, the load identification and decision module also compares the currently identified load status label with the historical load status label. When a jump in load status is detected, such as a direct switch from heavy load to light load, the load identification and decision module synchronously outputs a status switching indicator signal to the motor drive control module, triggering the drive execution submodule to perform fast response control.

[0048] like Figure 1 and Figure 2 As shown, the motor energy-saving method based on a load identification algorithm is implemented based on a motor energy-saving system based on a load identification algorithm according to any one of claims 1 to 8, comprising:

[0049] S1: The load status sensing module collects the motor's operating electrical parameters in real time, and generates standardized current timing data and voltage timing data after internal preprocessing;

[0050] S2: Receive the current timing data and voltage timing data through the dynamic load feature construction module, extract the dynamic change features of the load through a sliding time window, and generate load feature data;

[0051] S3: The load feature data is input into the pre-trained load identification model through the load identification decision module, and the current motor load state is identified through model inference and the corresponding load state label is output.

[0052] S4: Receive the load status tag through the motor drive control module, match the corresponding target control parameters according to the load status tag, and dynamically adjust the input voltage of the motor according to the target control parameters to achieve energy-saving operation of the motor.

[0053] The offline training process of the load recognition model in step S3 is as follows:

[0054] Pre-collect electrical parameter samples of the motor under various known load conditions, extract load feature data from each sample and label the corresponding load state. Use the labeled load feature data as a training set to input into the classification model to be trained. Through iterative optimization, the classification model learns and establishes the mapping relationship between the motor's operating electrical parameters and the load state. After training, the model parameters are fixed in the load identification decision module for use in the online inference stage.

[0055] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A motor energy-saving system based on a load identification algorithm, characterized in that, include: Load status sensing module, dynamic load characteristic construction module, load identification and decision module, and motor drive control module; A load status sensing module is connected to the motor power supply circuit to collect motor operating electrical parameters in real time. A dynamic load feature construction module is connected to the load state sensing module and is used to extract load feature data based on the motor operating electrical parameters. The load feature data includes multi-dimensional feature information characterizing the motor load change state over time. A load identification decision module, which is connected to the dynamic load feature construction module, is used to input the load feature data into a pre-trained load identification model to identify the current motor load state and output a load state label. A motor drive control module, which is connected to the load identification and decision module, is used to determine the target control parameters based on the load status label and adjust the input voltage or input frequency of the motor.

2. The motor energy-saving system based on a load identification algorithm according to claim 1, characterized in that, The dynamic load characteristic construction module includes: The time-series sampling submodule is used to continuously sample the motor operating electrical parameters within a preset detection period to obtain a time series of motor operating parameters; The feature extraction submodule is used to extract features from the time series of motor operating parameters to generate the load feature data.

3. The motor energy-saving system based on a load identification algorithm according to claim 1, characterized in that, The load identification decision module has a built-in load identification model trained on working condition samples. The load identification model is used to receive the load feature data and output the corresponding load status label. The load status label includes light load status, medium load status, heavy load status and overload status.

4. The motor energy-saving system based on a load identification algorithm according to claim 1, characterized in that, The motor drive control module includes: The parameter parsing submodule is used to query a preset parameter mapping table based on the load status label in order to determine the target control parameter; The drive execution submodule is used to generate drive control signals based on the target control parameters and to dynamically adjust the input voltage of the motor.

5. A motor energy-saving system based on a load identification algorithm according to claim 1, characterized in that, The drive execution submodule uses voltage regulation to adjust the motor input voltage. The target control parameter is a voltage regulation command, which includes voltage amplitude adjustment amount and voltage regulation rate.

6. A motor energy-saving system based on a load identification algorithm according to claim 1, characterized in that, The motor operating electrical parameters include the motor's three-phase current and three-phase voltage. The dynamic load feature construction module constructs the load feature data by extracting and combining features from the three-phase current and three-phase voltage.

7. A motor energy-saving system based on a load identification algorithm according to claim 1, characterized in that, The load identification decision module is also used to compare the identified load status label with the historical load status label. When the load status changes across levels, it outputs a status switching indicator signal. The motor drive control module responds to the status switching indicator signal to perform fast response control.

8. A motor energy-saving system based on a load identification algorithm according to claim 1, characterized in that, The load identification model is a classification model based on statistical learning or neural networks. It is obtained through offline training and the model parameters remain unchanged during system operation, outputting the load status label with fixed decision logic.

9. A motor energy-saving method based on a load identification algorithm, implemented based on a motor energy-saving system based on a load identification algorithm according to any one of claims 1 to 8, characterized in that, include: S1: The load status sensing module collects the motor's operating electrical parameters in real time, and generates standardized current timing data and voltage timing data after internal preprocessing; S2: Receive the current timing data and voltage timing data through the dynamic load feature construction module, extract the dynamic change features of the load through a sliding time window, and generate load feature data; S3: The load feature data is input into the pre-trained load recognition model through the load recognition decision module. The model then infers and identifies the current motor load state and outputs the corresponding load state label. S4: Receive the load status tag through the motor drive control module, match the corresponding target control parameters according to the load status tag, and dynamically adjust the input voltage of the motor according to the target control parameters to achieve energy-saving operation of the motor.

10. A motor energy-saving method based on a load identification algorithm according to claim 9, characterized in that, The offline training process of the load recognition model in step S3 includes: Samples of the motor's operating electrical parameters under various known load conditions are collected. Load feature data are extracted from the samples and labeled with corresponding load state labels. The labeled load feature data is used as a training set and input into the classification model to be trained. Through iterative optimization, the classification model learns the mapping relationship between the motor's operating electrical parameters and the load state.

Images