Energy-saving fan applied to inflatable model and control method of energy-saving fan

Abstract

The invention relates to the technical field of intelligent fan control, and particularly discloses an energy-saving fan applied to an inflatable model and a control method of the energy-saving fan applied to the inflatable model. And performing time sequence analysis on external wind speed and internal pressure data by using an artificial intelligence technology based on deep learning to capture a time sequence change mode of the external wind speed and the internal pressure, and further performing time sequence response analysis on the external wind speed and the internal pressure to excavate the associated influence of the external environment wind speed on the internal pressure of the inflation model. Therefore, the rotating speed of the fan is intelligently controlled by combining the external environment condition and the internal pressure state of the inflatable model, so that the stability of the inflatable model is maintained. In this way, energy consumption can be effectively reduced, meanwhile, the stability and safety of the inflation model under various environmental conditions are ensured, and therefore the energy efficiency of the inflation model is improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent fan control technology, and more specifically, to an energy-saving fan applied to an inflatable model and a control method thereof. Background Art

[0002] Inflatable models, as a structural form widely used in entertainment, advertising, exhibitions and emergency rescue, have always been the focus of attention in their design and use in terms of stability and energy efficiency. This type of model maintains its shape and volume through continuous air supply from an internal fan, ensuring that it can maintain a good display effect or function in various environments.

[0003] During the operation of the inflatable model, external environmental factors, especially changes in wind speed, have a direct and significant impact on the internal pressure state of the inflatable model. For example, in a strong wind environment, the external airflow directly acts on the surface of the model, which may cause a sharp change in internal pressure. If the fan control system cannot respond quickly, the model may be deformed or even damaged due to the large internal and external pressure difference. On the contrary, in weak or no wind conditions, excessive air supply is also an unnecessary energy consumption.

[0004] However, traditional fan control strategies mostly rely on simple sensor feedback, such as adjusting the fan speed by monitoring the internal pressure of the inflatable model and comparing it with the set pressure threshold. Although this method can maintain the basic shape of the model to a certain extent, it lacks adaptability to changes in the external environment. This may not only lead to energy waste, but also affect the structural stability and service life of the inflatable model due to the inability to adjust in time.

[0005] Therefore, an optimized energy-saving fan and a control method thereof for use in an inflatable model are desired. Summary of the invention

[0006] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides an energy-saving fan and a control method thereof applied to an inflatable model, which uses sensing technology to continuously monitor the external wind speed and internal pressure of the inflatable model, and uses artificial intelligence technology based on deep learning to perform time series analysis on the external wind speed and internal pressure data to capture the time series change pattern of the external wind speed and internal pressure, and then, by performing time series response analysis on the two, the correlation effect of the external environmental wind speed on the internal pressure of the inflatable model is excavated, so as to intelligently control the fan speed in combination with the external environmental conditions and internal pressure state of the inflatable model to maintain the stability of the inflatable model. In this way, energy consumption can be effectively reduced while ensuring the stability and safety of the inflatable model under various environmental conditions, thereby improving the energy efficiency of the inflatable model.

[0007] According to one aspect of the present application, a control method for an energy-saving fan applied to an inflatable model is provided, which includes:

[0008] A wind speed sensor deployed outside the inflatable model collects a time series data set of real-time wind speed values, and a pressure sensor deployed inside the inflatable model collects a time series data set of internal pressure values;

[0009] Transmitting the time series data set of the real-time wind speed value and the time series data set of the internal pressure value to the wind turbine central controller via a wireless transmission module;

[0010] At the wind turbine central controller, extracting time series features from the time series data set of the real-time wind speed value and the time series data set of the internal pressure value to obtain an external wind speed time series implicit correlation feature vector and a model internal pressure time series implicit correlation feature vector;

[0011] In the wind turbine central controller, a time series response analysis based on implicit correlation between features is performed on the external wind speed time series implicit correlation feature vector and the model internal pressure time series implicit correlation feature vector to obtain an external wind speed-internal pressure time series response coding vector;

[0012] In the fan central controller, a control signal for the fan speed is generated based on the external wind speed-internal pressure timing response coding vector, and the control signal is used to control the variable frequency drive of the energy-saving fan.

[0013] According to another aspect of the present application, there is provided an energy-saving fan for use in an inflatable model, comprising:

[0014] The inflatable model state monitoring module is used to collect a time series data set of real-time wind speed values ​​through a wind speed sensor deployed outside the inflatable model, and to collect a time series data set of internal pressure values ​​through a pressure sensor deployed inside the inflatable model;

[0015] A data transmission module, used to transmit the time series data set of the real-time wind speed value and the time series data set of the internal pressure value to the wind turbine central controller via a wireless transmission module;

[0016] A data time series feature extraction module is used to extract time series features from the time series data set of the real-time wind speed value and the time series data set of the internal pressure value in the wind turbine central controller to obtain an external wind speed time series implicit correlation feature vector and a model internal pressure time series implicit correlation feature vector;

[0017] A timing response analysis module is used to perform timing response analysis on the external wind speed timing implicit correlation feature vector and the model internal pressure timing implicit correlation feature vector based on implicit correlation between features in the wind turbine central controller to obtain an external wind speed-internal pressure timing response coding vector;

[0018] A control signal generating module is used to generate a control signal of the fan speed in the fan central controller based on the external wind speed-internal pressure timing response coding vector, and the control signal is used to control the variable frequency drive of the energy-saving fan.

[0019] Compared with the prior art, the energy-saving fan and control method thereof applied to the inflatable model provided by the present application adopts sensing technology to continuously monitor the external wind speed and internal pressure of the inflatable model, and uses artificial intelligence technology based on deep learning to perform time series analysis on the external wind speed and internal pressure data to capture the time series change pattern of the external wind speed and internal pressure, and then, by performing time series response analysis on the two, the correlation effect of the external environmental wind speed on the internal pressure of the inflatable model is excavated, so as to intelligently control the fan speed in combination with the external environmental conditions and internal pressure state of the inflatable model to maintain the stability of the inflatable model. In this way, energy consumption can be effectively reduced, while ensuring the stability and safety of the inflatable model under various environmental conditions, thereby improving the energy efficiency of the inflatable model. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0021] Figure 1 A schematic diagram of an application scenario of a control method of an energy-saving fan applied to an inflatable model according to an embodiment of the present application;

[0022] Figure 2 is a flow chart of a control method of an energy-saving fan applied to an inflatable model according to an embodiment of the present application;

[0023] Figure 3 A data flow diagram of a control method for an energy-saving fan applied to an inflatable model according to an embodiment of the present application;

[0024] Figure 4 is a flow chart of step S4 in the control method of an energy-saving fan applied to an inflatable model according to an embodiment of the present application;

[0025] Figure 5is a block diagram of an energy-saving fan applied to an inflatable model according to an embodiment of the present application;

[0026] In the figure: 1. Inflatable model; 2. Pressure sensor; 3. Wind speed sensor; 4. Fan central controller; 5. Frequency converter; 6. Energy-saving fan. DETAILED DESCRIPTION

[0027] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0028] In response to the technical problems described in the above background technology, this application proposes a control method for an energy-saving fan applied to an inflatable model, which uses sensing technology to continuously monitor the external wind speed and internal pressure of the inflatable model, and uses artificial intelligence technology based on deep learning to perform time series analysis on the external wind speed and internal pressure data to capture the time series change pattern of the external wind speed and internal pressure. Then, by performing time series response analysis on the two, the correlation effect of the external environmental wind speed on the internal pressure of the inflatable model is excavated, so as to intelligently control the fan speed in combination with the external environmental conditions and internal pressure state of the inflatable model to maintain the stability of the inflatable model. In this way, energy consumption can be effectively reduced, while ensuring the stability and safety of the inflatable model under various environmental conditions, thereby improving the energy efficiency of the inflatable model.

[0029] Figure 1 FIG. 1 is a schematic diagram of an application scenario of a control method for an energy-saving fan applied to an inflatable model according to an embodiment of the present application. Figure 1 As shown, in this application scenario, first, the real-time pressure value inside is collected by the pressure sensor 2 deployed inside the inflatable model 1, and the real-time wind speed value outside is collected by the wind speed sensor 3 deployed outside the inflatable model 1. Then, the time data set of the collected real-time pressure value and the time data set of the real-time wind speed value are input into the fan central controller 4 deployed with the fan speed control algorithm for data processing to generate a control signal for adjusting the fan speed, and the control signal is sent to the variable frequency driver 5 of the energy-saving fan 6 for fan speed control.

[0030] Figure 2 The present invention is a flowchart of a method for controlling an energy-saving fan applied to an inflatable model according to an embodiment of the present application. Figure 3 FIG. 1 is a data flow diagram of a control method for an energy-saving fan applied to an inflatable model according to an embodiment of the present application. Figure 2 and Figure 3As shown, the control method of the energy-saving fan applied to the inflatable model includes the following steps: S1, collecting a time series data set of real-time wind speed values ​​by a wind speed sensor deployed outside the inflatable model, and collecting a time series data set of internal pressure values ​​by a pressure sensor deployed inside the inflatable model; S2, transmitting the time series data set of the real-time wind speed values ​​and the time series data set of the internal pressure values ​​to the fan central controller through a wireless transmission module; S3, in the fan central controller, performing time series feature extraction on the time series data set of the real-time wind speed values ​​and the time series data set of the internal pressure values ​​to obtain an external wind speed time series implicit correlation feature vector and a model internal pressure time series implicit correlation feature vector; S4, in the fan central controller, performing time series response analysis based on implicit correlation between features on the external wind speed time series implicit correlation feature vector and the model internal pressure time series implicit correlation feature vector to obtain an external wind speed-internal pressure time series response coding vector; S5, in the fan central controller, generating a control signal for the fan speed based on the external wind speed-internal pressure time series response coding vector, and the control signal is used to control the variable frequency drive of the energy-saving fan.

[0031] In the control method of the energy-saving fan applied to the inflatable model, the step S1 collects a time series data set of real-time wind speed values ​​through a wind speed sensor deployed outside the inflatable model, and collects a time data set of internal pressure values ​​through a pressure sensor deployed inside the inflatable model. It should be understood that internal pressure is a key factor in maintaining the shape and structural integrity of the inflatable model. By monitoring the internal pressure, it can be ensured that the inflatable model is at an appropriate pressure level, neither over-inflation causing excessive stress on the material, nor under-pressure causing the model to collapse or become unstable. Changes in external wind speed will have a direct impact on the inflatable model. For example, under strong wind conditions, the inflatable model may be deformed or damaged due to external pressure. Therefore, by monitoring the external wind speed, the working state of the fan can be adjusted in time to ensure that the inflatable model can appropriately respond to changes in the external environment and maintain its shape and stability. For example, when the external wind speed is low and the internal pressure is stable, the fan speed can be reduced to reduce unnecessary energy consumption; on the contrary, when the external wind speed increases or the internal pressure decreases, the fan speed can be increased to ensure the normal operation of the inflatable model. This not only improves the response efficiency of fan control, but also achieves the goal of energy conservation and emission reduction.

[0032] It should be understood that when monitoring the internal and external conditions of an inflatable model, it is crucial to select the right sensor and ensure that it is deployed correctly. Specifically, when deploying wind speed sensors, the purpose is to accurately obtain the changes in wind force in the surrounding environment. This not only involves choosing the type of sensor, but also considering its specific installation location to ensure that the data obtained is representative. In outdoor environments, wind speed is a dynamic factor that varies with time and space. Therefore, in order to obtain the most realistic data reflecting the actual situation, a high-precision and durable wind speed sensor is required. Among them, ultrasonic or propeller anemometers are ideal because of their good response speed and accuracy. These sensors can provide stable measurements in all weather conditions and respond quickly to rapidly changing wind conditions.

[0033] Regarding the installation location of the wind speed sensor, it should be placed in a place that can represent the average wind speed of the entire inflatable model area. At the same time, in order to avoid local obstacles interfering with the measurement results, the sensor should be kept away from buildings or other objects that may affect the flow of air. In addition, if the inflatable model is located in an open area, several sensors can be arranged in multiple directions to further improve the reliability of the data through multi-point sampling. Doing so will help to understand the current wind speed conditions more accurately and provide an important basis for subsequent intelligent control.

[0034] Next, let's talk about the pressure monitoring inside the inflatable model. The pressure sensor used here must be sensitive enough to detect small pressure fluctuations. Given that the pressure inside the inflatable model is usually at a low level (relative to atmospheric pressure), a micro-differential pressure sensor can be used to perform this task. This type of sensor is designed for accurate measurements in the low pressure range and is very suitable for monitoring the pressure status inside the inflatable model. In addition, since there may be different pressure distributions inside the inflatable model, when choosing the installation location of the sensor, the sensor should be installed in a position that can reflect the overall pressure level while avoiding direct exposure to environmental factors that may cause sudden changes in pressure. For example, the sensor can be installed in the central area of ​​the inflatable model or near the connection point of the supporting structure, which usually provides more stable and representative pressure readings.

[0035] After ensuring the type and installation location of the sensor, the next step is to collect data. In order to ensure the quality of the data, a reasonable sampling frequency must be set. In dynamic environments such as outdoor exhibitions, a higher sampling rate (such as several times per second) is necessary, because the external conditions change rapidly at this time, and frequent data updates help the control system respond faster. In relatively stable indoor display scenes, it is sufficient to maintain a lower sampling frequency, which can meet monitoring needs without generating too much redundant data. It is worth noting that in either case, the data collection of the wind speed sensor and the pressure sensor should be synchronized, so that the values ​​of both can be obtained at the same time, which is conducive to the subsequent analysis of the correlation between the two.

[0036] In addition to setting an appropriate sampling frequency, the raw data also needs to be preliminarily processed. Since the data from the sensor often contains noise and other forms of errors, it may lead to wrong judgments if it is not cleaned up. Therefore, the data can be preprocessed at the sensor end, including removing unnecessary noise, calibrating possible deviations, and other measures. In addition, outliers should also be filtered to prevent erroneous readings in extreme cases from affecting the decision-making process of the control system. By preprocessing the data, the accuracy of subsequent analysis can be greatly improved, and the workload of the central controller can also be reduced.

[0037] In the above-mentioned control method of the energy-saving fan applied to the inflatable model, the step S2 transmits the time series data set of the real-time wind speed value and the time series data set of the internal pressure value to the fan central controller through the wireless transmission module. It should be understood that in order to achieve efficient transmission and processing of data, the present application further transmits the time series data set of the real-time wind speed value and the time series data set of the internal pressure value to the fan central controller through the wireless transmission module, so as to use the data processing algorithm deployed by the fan central controller to perform data analysis on the collected wind speed data and pressure data, and realize intelligent control of the fan speed. In addition, in the technical solution of the present application, the use of a wireless transmission module for data transmission can effectively reduce the wiring cost and maintenance difficulty, especially in outdoor or large venues, avoiding the trouble and extra expenses caused by laying long-distance cables. In addition, modern wireless transmission technology can provide sufficiently low latency to support real-time data transmission, so as to ensure that the fan central controller can quickly receive the latest wind speed and pressure information, so as to respond in time.

[0038] Specifically, in the specific implementation process of using the wireless transmission module to transmit the time series data set of the real-time wind speed value and the time series data set of the internal pressure value to the central controller of the wind turbine, it includes but is not limited to selecting a suitable wireless communication technology, ensuring the security and reliability of data transmission, optimizing the network configuration to support large-scale deployment, and coping with various challenges that may be encountered. Accordingly, the following is a comprehensive description of this process:

[0039] First of all, in order to ensure the effectiveness and efficiency of data transmission, choosing the appropriate wireless communication technology is the key first step. Considering the characteristics of the application scenario, that is, the need to transmit data over long distances in a relatively open space, and to ensure low latency and high stability, several common wireless communication protocols have become candidates: Wi-Fi, LoRa (Long Range), Zigbee and Bluetooth Low Energy (BLE). Each protocol has its own unique advantages and disadvantages and is suitable for different types of project requirements. For this application, a solution that can provide stable connections over a large range and has low power consumption characteristics is required. Therefore, after comprehensively evaluating factors such as transmission distance, bandwidth requirements, and energy consumption levels, LoRa became the first choice.

[0040] LoRa technology is known for its long-distance coverage capabilities and can maintain good signal quality even in complex terrain conditions. The operating frequency of LoRa technology is usually in the unlicensed band, which means that it can be used without additional spectrum resources. In addition, LoRa devices consume very little energy and are very suitable for application scenarios that have high requirements for battery life. More importantly, the LoRa network architecture supports a star topology, which makes it easy to expand to multiple nodes to meet the needs of large inflatable model sites. By installing LoRa gateways at appropriate locations around the inflatable model, a stable wireless network environment can be established, allowing all deployed sensors to access it and upload data to the wind turbine central controller in real time.

[0041] However, after selecting the appropriate wireless communication technology, it is also necessary to ensure that the transmitted data has a high degree of security and integrity. After all, these data are directly related to the normal operation of the wind turbine control system, and any tampering or loss may lead to serious consequences. To this end, a variety of security measures are adopted to protect data during wireless transmission. On the one hand, encryption algorithms are used to encrypt the sent data packets to prevent unauthorized third parties from stealing sensitive information; on the other hand, the legitimacy of the identity of each device is confirmed through an authentication mechanism, and only authenticated devices can join the network and participate in communication. At the same time, access rights rules can also be set to restrict specific devices to perform only certain operations, thereby further enhancing the security of the system.

[0042] In addition to basic security guarantees, it is also necessary to consider how to optimize the network configuration to support large-scale deployment. When there are a large number of inflatable models or they occupy a large area, multiple sensors may be involved in working at the same time. In this case, if network resources cannot be effectively managed, congestion can easily occur, which in turn affects the overall performance. In order to solve this problem, some intelligent scheduling algorithms can be introduced, such as priority-based data queuing mechanisms or methods that dynamically adjust channel occupancy rates to ensure that important data is processed first, while unimportant data can be transmitted during idle periods. In addition, the concept of edge computing can be used to deploy small servers or gateway devices close to sensors to complete some data analysis tasks in advance, reduce the pressure on the central server, and improve response speed.

[0043] Of course, it is inevitable to encounter various emergencies during the actual deployment process. For example, due to the influence of natural environmental factors (such as bad weather), the wireless signal may be attenuated or even interrupted. In response to this situation, it is necessary to formulate emergency plans in advance, such as adding backup routing paths or using self-healing network technology. Once the main link has problems, the system can automatically switch to the backup link to continue working. In addition, considering the risk of hardware failure during long-term operation, a comprehensive maintenance system should be established to regularly check the status of equipment and replace aging components in a timely manner to ensure the continued stable operation of the entire system.

[0044] To summarize, the specific process of data transmission includes: First, the data collected from the sensor will be preliminarily processed locally to remove noise, calibration errors, etc., and then packaged into data frames according to the preset format. Each data frame contains not only the actual measured values ​​of wind speed and pressure, but also metadata such as timestamps and device identifiers for subsequent parsing. Subsequently, these data frames are sent out through the selected wireless communication protocol, and finally reach the wind turbine central controller after a series of intermediate links (such as gateway forwarding). Here, the central controller will decode and parse the received data and store it in the database for further analysis. It is worth noting that in order to adapt to the requirements of different application scenarios, the central controller should also have a flexible data receiving interface that can be compatible with multiple types of input sources, whether from a single sensor or multiple distributed nodes.

[0045] In the control method of the energy-saving fan applied to the inflatable model, in the step S3, in the central controller of the fan, the time series data set of the real-time wind speed value and the time series data set of the internal pressure value are subjected to time series feature extraction to obtain the external wind speed time series implicit correlation feature vector and the model internal pressure time series implicit correlation feature vector. In a specific example of the present application, the time series data set of the real-time wind speed value and the time series data set of the internal pressure value are input into the time series encoder based on the LSTM-RNN hybrid model to obtain the external wind speed time series implicit correlation feature vector and the model internal pressure time series implicit correlation feature vector. Here, considering that the external wind speed and internal pressure of the inflatable model are dynamic data that change with time and have a high degree of time series characteristics, therefore, in order to fully capture the time series change pattern of the wind speed data and the pressure data, the present application adopts the LSTM-RNN hybrid model to construct a time series encoder to extract the time series features of the time series data set of the real-time wind speed value and the time series data set of the internal pressure value, respectively, to obtain the external wind speed time series implicit correlation feature vector and the model internal pressure time series implicit correlation feature vector. Those skilled in the art should know that both the LSTM (Long Short-Term Memory Network) model and the RNN (Recurrent Neural Network) model are deep learning models that are good at processing time series data. Among them, the RNN model can effectively capture the short-term dependencies in the time series data through its internal cyclic connection structure, but because it faces the problem of gradient vanishing or gradient exploding in practical applications, the RNN model often does not work well when processing long sequence data. The LSTM model can better capture and utilize the long-term dependency information in the sequence data by introducing a gating mechanism, avoiding the gradient vanishing problem in the traditional RNN model. Therefore, in the technical solution of the present application, in order to comprehensively utilize the advantages of the LSTM model and the RNN model, a LSTM-RNN hybrid model is constructed, in order to capture short-term dynamic changes and retain long-term dependency information in the process of extracting the time series features of wind speed data and pressure data, thereby providing a more accurate data basis for subsequent fan speed control.

[0046] In the above-mentioned control method of the energy-saving fan applied to the inflatable model, in step S4, at the fan central controller, the external wind speed time series implicitly associated characteristic vector and the model internal pressure time series implicitly associated characteristic vector are subjected to a time series response analysis based on the implicit association between the characteristics to obtain the external wind speed-internal pressure time series response coding vector. That is, considering that for the inflatable model, the dynamic change between the external wind speed and the internal pressure is not an isolated event, but there is a certain correlation. Therefore, the present application further performs a time series response analysis on the external wind speed time series implicitly associated characteristic vector and the model internal pressure time series implicitly associated characteristic vector to reveal the functional relationship between the external wind speed and the internal pressure, so as to fully understand the time series change pattern of the internal pressure of the inflatable model under specific external wind speed conditions, so as to timely adjust the fan speed to adapt to this pressure change, and avoid the inflatable model from being damaged or over-inflated due to sudden pressure changes.

[0047] Figure 4 FIG. 4 is a flow chart of step S4 in a method for controlling an energy-saving fan for an inflatable model according to an embodiment of the present application. Figure 4 As shown, the step S4 includes: S41, inputting the implicit correlation feature vector of the external wind speed time series and the implicit correlation feature vector of the model internal pressure time series into a collaborative feature extraction network to obtain an implicit collaborative coding vector between the external wind speed-internal pressure time series features; S42, based on the implicit collaborative coding vector between the external wind speed-internal pressure time series features, performing feature modulation interactive fusion on the implicit correlation feature vector of the external wind speed time series and the implicit correlation feature vector of the model internal pressure time series to obtain the external wind speed-internal pressure time series response coding vector.

[0048] In a specific example of the present application, the collaborative feature extraction network includes three parallel feature interaction layers, a multi-level interactive feature cascade layer, a point convolution layer, and an activation layer based on a Leaky ReLU function. Based on this, the calculation process of step S41 can be expressed as:

[0049]

[0050] Wherein, V1 represents the implicit correlation feature vector of the external wind speed time series, V2 represents the implicit correlation feature vector of the model internal pressure time series, represents point-by-point addition, ⊙ represents point-by-point multiplication, represents point-by-point subtraction, concat(·;·;·) represents feature concatenation, Conv 1×1 represents a 1×1 convolution operation, Leaky ReLU represents the Leaky ReLU activation function, and V J Represents the implicit collaborative coding vector between the external wind speed and internal pressure time series features.

[0051] That is, firstly, the collaborative feature extraction network is used to perform multi-level interaction, point convolution, nonlinear activation and other operations on the implicit correlation feature vector of the external wind speed time series and the implicit correlation feature vector of the internal pressure time series of the model, so as to capture the deep correlation between the two and mine the time series collaborative feature representation between the two.

[0052] In a specific example of the present application, the step S42 includes: first, writing the implicit collaborative coding vector between the external wind speed and internal pressure time series features into a dynamic memory unit to obtain a dynamic key vector, which is expressed as:

[0053] m u =M(V J )

[0054] Where M(·) represents a 1×1 convolution operation, m u Represents a dynamic key vector.

[0055] Here, the dynamic memory unit is a special type of memory structure, which works similarly to the memory unit in a recursive neural network. It can dynamically adjust its own memory content and internal state according to the current input features to better adapt to the contextual requirements of the current scene.

[0056] Next, based on the dynamic key vector, the external wind speed time series implicit association feature vector and the model internal pressure time series implicit association feature vector are respectively subjected to feature attention modulation optimization to obtain the optimized external wind speed time series implicit association feature vector and the optimized model internal pressure time series implicit association feature vector. More specifically, first, the dynamic key vector is extracted from the dynamic memory unit, and the external wind speed time series implicit association feature vector and the dynamic key vector are input into the feature attention modulation module based on the first converter structure to obtain the optimized external wind speed time series implicit association feature vector, which is expressed as:

[0057]

[0058]

[0059] Among them, W 1q and W 1v denote the first query embedding matrix and the first value embedding matrix respectively, V 1q and V 1v denote the first query vector and the first value vector, respectively, and b 1q and b 1vThey represent different bias terms respectively, softmax(·) represents the normalized exponential function, S represents the characteristic scale value of the dynamic key vector, and V'1 represents the implicit correlation characteristic vector of the external wind speed time series after optimization.

[0060] Then, the dynamic key vector is extracted from the dynamic memory unit, and the model internal pressure time series implicit association feature vector and the dynamic key vector are input into the feature attention modulation module based on the second converter structure to obtain the optimized model internal pressure time series implicit association feature vector, which is expressed as:

[0061]

[0062]

[0063] Among them, W 2q and W 2v denote the second query embedding matrix and the second value embedding matrix, V 2q and V 2v denote the second query vector and the second value vector, respectively, and b 2q and b 2v They represent different bias terms respectively, and V'2 represents the implicit correlation eigenvector of the internal pressure time series of the optimized model.

[0064] That is, respectively construct a query vector and a value vector for the external wind speed time series implicit association feature vector, and a query vector and a value vector for the model internal pressure time series implicit association feature vector, and utilize the self-attention mechanism in the converter structure to perform query attention-based feature modulation on the external wind speed time series implicit association feature vector and the model internal pressure time series implicit association feature vector based on the dynamic key vector (i.e., the time series collaborative feature between the external wind speed and the internal pressure). This is equivalent to using the dynamic key vector as a common attention space to enhance the expression of key information in the original features and reduce the influence of noise and irrelevant information. At the same time, it also makes the optimized external wind speed time series implicit association feature vector and the model internal pressure time series implicit association feature vector in an aligned semantic space to facilitate the subsequent feature interaction and fusion processing between the two.

[0065] Finally, the optimized external wind speed time series implicit correlation feature vector and the optimized model internal pressure time series implicit correlation feature vector are fused to obtain the external wind speed-internal pressure time series response encoding vector. More specifically, the optimized external wind speed time series implicit correlation feature vector and the optimized model internal pressure time series implicit correlation feature vector are input into the characteristic linear interaction network to obtain the external wind speed-internal pressure time series response encoding vector, which is expressed as:

[0066] V=γV'1+(1-γ)V'2

[0067] Among them, γ is the fusion weight parameter, and V represents the external wind speed-internal pressure time series response encoding vector.

[0068] That is, a characteristic linear interactive network is used to interactively fuse the implicitly associated characteristic vectors of the optimized external wind speed time series and the implicitly associated characteristic vectors of the optimized model internal pressure time series. That is, a weighted fusion of the wind speed time series change characteristics and the pressure time series change characteristics is performed through linear transformation to generate an external wind speed-internal pressure time series response encoding vector, thereby realizing a comprehensive response analysis of the inflation model under different external wind speed and internal pressure conditions, and providing more accurate and comprehensive decision support for the subsequent intelligent control of the fan speed.

[0069] In the control method of the energy-saving fan applied to the inflatable model, in step S5, the fan central controller generates a control signal of the fan speed based on the external wind speed-internal pressure time series response coding vector, and the control signal is used to control the variable frequency drive of the energy-saving fan. In a specific example of the present application, step S5 includes: inputting the external wind speed-internal pressure time series response coding vector into a decoder-based wind speed optimizer to obtain a decoded value of the optimal fan speed; based on the comparison between the decoded value of the optimal fan speed and the current fan speed value, generating the control signal.

[0070] Specifically, after receiving the external wind speed-internal pressure time series response encoding vector, the decoder structure in the wind speed optimizer learns and simulates the complex dynamic relationship between the external wind speed and the internal pressure by extracting features layer by layer, and based on this, predicts the optimal speed that the fan should reach under the current external wind speed and internal pressure conditions. In an embodiment of the present application, the decoder adopts a multi-layer perceptron (MLP) model.

[0071] Finally, according to the difference between the optimal speed decoded value provided by the decoder and the current fan speed value, the fan central controller further calculates the corresponding adjustment amount through the control algorithm, generates the corresponding control signal, and adjusts the fan speed through the variable frequency drive to achieve energy saving and efficiency optimization. In this way, it can not only ensure the stable operation of the fan under various wind speed and pressure conditions, but also effectively extend the service life of the fan and reduce the additional energy consumption caused by frequent speed adjustment.

[0072] Preferably, inputting the external wind speed-internal pressure time series response encoding vector into a decoder-based wind speed optimizer to obtain a decoded value of the optimal wind turbine speed comprises:

[0073] First, the absolute value sum and the square root of the square sum of all eigenvalues ​​of the external wind speed-internal pressure time series response encoding vector are calculated to obtain the first and second external wind speed-internal pressure time series response encoding spatial structure values, that is:

[0074] w1=∑ i |v i |

[0075]

[0076] Among them, v i is the eigenvalue of the ith position in the external wind speed-internal pressure time series response coding vector, w1 is the first external wind speed-internal pressure time series response coding space structure value, and w2 is the second external wind speed-internal pressure time series response coding space structure value;

[0077] Next, determining the number L of eigenvalues ​​of the external wind speed-internal pressure time series response encoding vector;

[0078] Then, for each eigenvalue of the external wind speed-internal pressure time series response encoding vector, the first external wind speed-internal pressure time series response encoding spatial structure value minus the product of the eigenvalue and the number of eigenvalues ​​is calculated to obtain the first external wind speed-internal pressure time series response encoding long-range dependency value, that is:

[0079] x i =w1-v i ×L

[0080] Among them, x i For the v i The corresponding first external wind speed-internal pressure time series response encodes the long-range dependency value;

[0081] At the same time, the second external wind speed-internal pressure time series response encoding spatial structure value is calculated by multiplying the square root of the number of eigenvalues ​​by the product of the eigenvalues ​​minus the second external wind speed-internal pressure time series response encoding long-range dependency value, that is:

[0082]

[0083] Among them, y i For the v i The corresponding second external wind speed-internal pressure time series response encodes the long-range dependency value;

[0084] Secondly, the exponential value calculated by taking the first external wind speed-internal pressure time series response encoding long-range dependency value as the exponent of the natural constant and the inverse of the second external wind speed-internal pressure time series response encoding long-range dependency value are weighted summed to obtain the optimized eigenvalue corresponding to each eigenvalue, that is:

[0085]

[0086] Among them, α and β represent different weighting parameters, v' i Indicates the v i The corresponding optimized eigenvalue;

[0087] Finally, the optimized external wind speed-internal pressure time series response encoding vector composed of the optimized characteristic values ​​is input into a decoder-based wind speed optimizer to obtain a decoded value of the optimal wind turbine speed.

[0088] Here, considering that the distribution of the external wind speed time series implicit correlation feature vector and the model internal pressure time series implicit correlation feature vector represents the multi-scale time series correlation characteristics of the real-time wind speed value and the internal pressure value of the inflation model, after the attention interaction between the features, the external wind speed-internal pressure time series response encoding vector will also have significant attention interaction distribution spatial structure differences due to the dynamic memory differences in the time series correlation feature distribution caused by the differences in the time series distribution of the source data, affecting the convergence consistency of the decoder, and thus affecting the accuracy of the decoded value of the optimal wind speed obtained by the wind speed optimizer based on the decoder.

[0089] Based on this, in view of the possible spatial structure loss of the feature set of the external wind speed-internal pressure timing response coding vector in the high-dimensional space, which causes the weight matrix of the decoder to implicitly infer the spatial structure information based on the features, resulting in inconsistent convergence. A long-distance feature dependency relationship is established based on the spatial structure representation of the external wind speed-internal pressure timing response coding vector based on the overall feature scale of the external wind speed-internal pressure timing response coding vector to establish the feature local connectivity of the external wind speed-internal pressure timing response coding vector, and the spatial ambiguous information of the object feature value is captured by predicting the unstructured feature value points of the external wind speed-internal pressure timing response coding vector, thereby improving the spatial inductive deviation perception ability of the feature set of the external wind speed-internal pressure timing response coding vector, improving the convergence consistency of the decoder, and improving the accuracy of the decoded value of the optimal wind turbine speed obtained by the wind speed optimizer based on the decoder input of the external wind speed-internal pressure timing response coding vector.

[0090] Furthermore, the present application also provides an energy-saving fan for use in an inflatable model.

[0091] Figure 5 FIG. 1 is a block diagram of an energy-saving fan applied to an inflatable model according to an embodiment of the present application. Figure 5As shown, according to the embodiment of the present application, the energy-saving fan 100 applied to the inflatable model includes: an inflatable model state monitoring module 110, which is used to collect a time series data set of real-time wind speed values ​​through a wind speed sensor deployed on the outside of the inflatable model, and collect a time data set of internal pressure values ​​through a pressure sensor deployed inside the inflatable model; a data transmission module 120, which is used to transmit the time series data set of the real-time wind speed value and the time series data set of the internal pressure value to the fan central controller through a wireless transmission module; a data time series feature extraction module 130, which is used to extract the time series data set of the real-time wind speed value and the time series data set of the internal pressure value in the fan central controller. Performing time series feature extraction to obtain an external wind speed time series implicit correlation feature vector and a model internal pressure time series implicit correlation feature vector; a time series response analysis module 140, used to perform a time series response analysis based on implicit correlation between features on the external wind speed time series implicit correlation feature vector and the model internal pressure time series implicit correlation feature vector in the wind turbine central controller to obtain an external wind speed-internal pressure time series response coding vector; a control signal generation module 150, used to generate a control signal for the fan speed in the wind turbine central controller based on the external wind speed-internal pressure time series response coding vector, and the control signal is used to control the variable frequency drive of the energy-saving fan.

[0092] Here, those skilled in the art will appreciate that the specific operations of the various modules in the energy-saving fan for the inflatable model have been described in the above reference. Figures 1 to 4 The control method of the energy-saving fan applied to the inflation model has been described in detail, and therefore, its repeated description will be omitted.

[0093] The basic principle of the present invention is described above in conjunction with specific embodiments. However, it should be pointed out that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. must be possessed by each embodiment of the present invention. In addition, the specific details of the above embodiments are only for the purpose of illustration and facilitation of understanding, rather than limitation, and the above details do not limit the present invention to being implemented by adopting the above specific details.

Claims

1. A control method for an energy-saving fan applied to an inflatable model, characterized in that: include: A wind speed sensor deployed outside the inflatable model collects a time series data set of real-time wind speed values, and a pressure sensor deployed inside the inflatable model collects a time series data set of internal pressure values; Transmitting the time series data set of the real-time wind speed value and the time series data set of the internal pressure value to the wind turbine central controller via a wireless transmission module; At the wind turbine central controller, extracting time series features from the time series data set of the real-time wind speed value and the time series data set of the internal pressure value to obtain an external wind speed time series implicit correlation feature vector and a model internal pressure time series implicit correlation feature vector; In the wind turbine central controller, a time series response analysis based on implicit correlation between features is performed on the external wind speed time series implicit correlation feature vector and the model internal pressure time series implicit correlation feature vector to obtain an external wind speed-internal pressure time series response coding vector; In the fan central controller, a control signal for the fan speed is generated based on the external wind speed-internal pressure timing response coding vector, and the control signal is used to control the variable frequency drive of the energy-saving fan.

2. The control method of the energy-saving fan applied to the inflatable model according to claim 1 is characterized in that: Performing time series feature extraction on the time series data set of the real-time wind speed value and the time series data set of the internal pressure value to obtain an external wind speed time series implicit correlation feature vector and a model internal pressure time series implicit correlation feature vector, including: The time series data set of the real-time wind speed value and the time series data set of the internal pressure value are input into a time series encoder based on the LSTM-RNN hybrid model to obtain the external wind speed time series implicit correlation feature vector and the model internal pressure time series implicit correlation feature vector.

3. The control method of the energy-saving fan applied to the inflatable model according to claim 2 is characterized in that: Performing a time series response analysis based on implicit correlation between features on the external wind speed time series implicit correlation feature vector and the model internal pressure time series implicit correlation feature vector to obtain an external wind speed-internal pressure time series response coding vector, including: Inputting the implicit correlation feature vector of the external wind speed time series and the implicit correlation feature vector of the model internal pressure time series into a collaborative feature extraction network to obtain an implicit collaborative coding vector between the external wind speed and internal pressure time series features; Based on the implicit collaborative coding vector between the external wind speed-internal pressure time series features, the external wind speed time series implicit correlation feature vector and the model internal pressure time series implicit correlation feature vector are feature modulated and interactively fused to obtain the external wind speed-internal pressure time series response coding vector.

4. The control method of the energy-saving fan applied to the inflatable model according to claim 3 is characterized in that: The collaborative feature extraction network includes three parallel feature interaction layers, a multi-level interactive feature cascade layer, a point convolution layer and an activation layer based on the Leaky ReLU function.

5. The control method of the energy-saving fan applied to the inflatable model according to claim 4 is characterized in that: Based on the implicit collaborative coding vector between the external wind speed-internal pressure time series features, the external wind speed time series implicit correlation feature vector and the model internal pressure time series implicit correlation feature vector are subjected to feature modulation interactive fusion to obtain the external wind speed-internal pressure time series response coding vector, including: Writing the implicit collaborative coding vector between the external wind speed and internal pressure time series features into a dynamic memory unit to obtain a dynamic key vector; Based on the dynamic key vector, the external wind speed time series implicit correlation feature vector and the model internal pressure time series implicit correlation feature vector are respectively subjected to feature attention modulation optimization to obtain an optimized external wind speed time series implicit correlation feature vector and an optimized model internal pressure time series implicit correlation feature vector; The optimized external wind speed time series implicit correlation feature vector and the optimized model internal pressure time series implicit correlation feature vector are fused to obtain the external wind speed-internal pressure time series response encoding vector.

6. The control method of the energy-saving fan applied to the inflatable model according to claim 5 is characterized in that: Based on the dynamic key vector, the external wind speed time series implicit correlation feature vector and the model internal pressure time series implicit correlation feature vector are respectively subjected to feature attention modulation optimization to obtain an optimized external wind speed time series implicit correlation feature vector and an optimized model internal pressure time series implicit correlation feature vector, including: Extract the dynamic key vector from the dynamic memory unit, and input the external wind speed time series implicit association feature vector and the dynamic key vector into a feature attention modulation module based on a first converter structure to obtain the optimized external wind speed time series implicit association feature vector; The dynamic key vector is extracted from the dynamic memory unit, and the model internal pressure time series implicit association feature vector and the dynamic key vector are input into a feature attention modulation module based on a second converter structure to obtain the optimized model internal pressure time series implicit association feature vector.

7. The control method of the energy-saving fan applied to the inflatable model according to claim 6 is characterized in that: The optimized external wind speed time series implicit correlation feature vector and the optimized model internal pressure time series implicit correlation feature vector are integrated to obtain the external wind speed-internal pressure time series response encoding vector, including: The optimized external wind speed time series implicitly associated feature vector and the optimized model internal pressure time series implicitly associated feature vector are input into a characteristic linear interaction network to obtain the external wind speed-internal pressure time series response encoding vector.

8. The control method of the energy-saving fan applied to the inflatable model according to claim 7 is characterized in that: Based on the external wind speed-internal pressure time series response encoding vector, a control signal of the fan speed is generated, and the control signal is used to control the variable frequency drive of the energy-saving fan, including: Inputting the external wind speed-internal pressure time series response encoding vector into a decoder-based wind speed optimizer to obtain a decoded value of the optimal wind turbine speed; The control signal is generated based on a comparison between the decoded value of the optimal fan speed and a current fan speed value.

9. An energy-saving fan for an inflatable model, characterized in that: include: The inflatable model state monitoring module is used to collect a time series data set of real-time wind speed values ​​through a wind speed sensor deployed outside the inflatable model, and to collect a time series data set of internal pressure values ​​through a pressure sensor deployed inside the inflatable model; A data transmission module, used to transmit the time series data set of the real-time wind speed value and the time series data set of the internal pressure value to the wind turbine central controller via a wireless transmission module; A data time series feature extraction module is used to extract time series features from the time series data set of the real-time wind speed value and the time series data set of the internal pressure value in the wind turbine central controller to obtain an external wind speed time series implicit correlation feature vector and a model internal pressure time series implicit correlation feature vector; A timing response analysis module is used to perform timing response analysis on the external wind speed timing implicit correlation feature vector and the model internal pressure timing implicit correlation feature vector based on implicit correlation between features in the wind turbine central controller to obtain an external wind speed-internal pressure timing response coding vector; A control signal generating module is used to generate a control signal of the fan speed in the fan central controller based on the external wind speed-internal pressure timing response coding vector, and the control signal is used to control the variable frequency drive of the energy-saving fan.

Images