An underground parking lot fan prediction control method, system and storage medium

Abstract

This application relates to the field of intelligent control technology, and particularly to a method, system and storage medium for predicting and controlling a fan in an underground parking lot. Based on the vehicle entry and exit data of the underground parking lot, the vehicle flow data of each area, and a dynamic prediction model, this application determines the predicted ventilation strategies corresponding to multiple preset future time periods in the underground parking lot; based on the temperature and humidity change data of the underground parking lot, the air quality change data of each area, and a comfort feedback model, it optimizes the ventilation strategy of the target area in the underground parking lot to generate the target ventilation strategies corresponding to multiple preset future time periods in the underground parking lot. By predicting the ventilation requirements in preset future time periods and perceiving environmental changes in real time, this application dynamically adjusts the ventilation strategy, which can reduce energy consumption while improving the comfort of drivers.

Description

Technical Field

[0001] The present application relates to the technical field of intelligent control, and particularly to a method, a system, and a storage medium for predicting and controlling a fan in an underground parking lot. Background Art

[0002] With the acceleration of the urbanization pace, underground parking lots have gradually evolved into one of the key elements of urban infrastructure. Although they play an indispensable role in providing convenient parking services, due to their enclosed structural characteristics, underground parking lots often need to implement all-weather ventilation measures to ensure the circulation of internal air. This requirement, to a certain extent, has led to a significant consumption of energy, especially during periods when the vehicle usage frequency is low.

[0003] The operation mode of traditional ventilation systems is usually limited by a fixed schedule or a simple induction mechanism, lacking the flexibility to achieve precise control of ventilation strategies. The deficiencies of such systems are that they cannot effectively avoid the occurrence of insufficient or excessive ventilation, thus having a negative impact on air quality and the comfort of drivers. Therefore, there is an urgent need for intelligent and refined control of the ventilation systems in underground parking lots. Summary of the Invention

[0004] In view of this, the embodiments of the present application at least provide a method, a system, and a storage medium for predicting and controlling a fan in an underground parking lot, which can reduce energy consumption while improving the comfort of drivers.

[0005] The present application mainly includes the following aspects:

[0006] In a first aspect, the embodiments of the present application provide a method for predicting and controlling a fan in an underground parking lot, which is applied to a ventilation optimization device. The method includes:

[0007] Obtain the vehicle entry and exit data of the underground parking lot, the vehicle flow data of each area, the temperature and humidity change data of the underground parking lot collected in real time, and the air quality change data of each area collected by a data acquisition device within a first preset historical time period;

[0008] Based on the vehicle entry and exit data of the underground parking lot, the vehicle flow data of each area, and a dynamic prediction model, determine the predicted ventilation strategies corresponding to the underground parking lot in multiple preset future time periods; the predicted ventilation strategies include the ventilation strategies for each area;

[0009] Based on the temperature and humidity change data of the underground parking lot, the air quality change data of each area, and a comfort feedback model, optimize the ventilation strategy of the target area of the underground parking lot, generate the target ventilation strategies corresponding to the underground parking lot in multiple preset future time periods, and send the target ventilation strategies to a fan control device.

[0010] In a second aspect, an embodiment of the present application further provides an underground parking lot fan prediction control system, which includes a data acquisition device, a ventilation optimization device, and a fan control device. The ventilation optimization device is communicatively connected to the data acquisition device and the fan control device respectively; wherein,

[0011] The data acquisition device is configured to collect the vehicle entry and exit data, temperature and humidity change data at each moment, as well as the vehicle flow data and air quality change data in each area of the underground parking lot;

[0012] The ventilation optimization device is configured to determine the predicted ventilation strategies corresponding to the underground parking lot in multiple preset future time periods based on the vehicle entry and exit data, vehicle flow data in each area, and a dynamic prediction model of the underground parking lot in a first preset historical time period. The predicted ventilation strategies include the ventilation strategies for each area; based on the real-time temperature and humidity change data, air quality change data in each area, and a comfort feedback model of the underground parking lot, optimize the ventilation strategy of the target area of the underground parking lot, generate the target ventilation strategies corresponding to the underground parking lot in multiple preset future time periods, and send the target ventilation strategies to the fan control device;

[0013] The fan control device is configured to control the ventilation equipment in each area of the underground parking lot to ventilate according to the ventilation control instructions corresponding to the target ventilation strategy.

[0014] In a third aspect, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is run by a processor, it executes the steps of the underground parking lot fan prediction control method described in the first aspect or any possible implementation manner in the first aspect.

[0015] The underground parking lot fan prediction control method, system, and storage medium provided by the embodiments of the present application dynamically adjust the ventilation strategy by predicting the ventilation demand in a preset future time period and real-time sensing the environmental changes. Compared with the operation mode of traditional ventilation systems, which is usually limited by a fixed schedule or a simple induction mechanism and lacks flexibility to achieve precise control of ventilation strategies, the disadvantage of such systems is that they cannot effectively avoid the occurrence of insufficient or excessive ventilation, thereby having a negative impact on air quality and the comfort of drivers. In contrast, the embodiments of the present application can reduce energy consumption while improving the comfort of drivers.

[0016] To make the above objects, features, and advantages of the present application more obvious and understandable, the following specially provides preferred embodiments and, in conjunction with the accompanying drawings, makes a detailed description as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] To more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present application and should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0018] Figure 1 Shows a flowchart of a method for predicting and controlling a fan in an underground parking lot provided by an embodiment of the present application;

[0019] Figure 2 Shows a schematic diagram of the overall area of an underground parking lot and the installation positions of sensors provided by an embodiment of the present application;

[0020] Figure 3 Shows a schematic diagram of any area in an underground parking lot and the installation positions of sensors provided by an embodiment of the present application;

[0021] Figure 4 Shows a schematic diagram of the structure of a fan prediction control system for an underground parking lot provided by an embodiment of the present application;

[0022] Figure 5 Shows a schematic diagram of the working process of a fan prediction control system for an underground parking lot provided by an embodiment of the present application. Detailed implementation manners

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the drawings in the present application are only for the purpose of illustration and description and are not used to limit the protection scope of the present application. Additionally, it should be understood that the schematic drawings are not drawn to scale. The flowcharts used in the present application show the operations implemented according to some embodiments of the present application. It should be understood that the operations in the flowcharts may not be implemented in sequence, and steps without logical context relationships may be reversed or implemented simultaneously. In addition, those skilled in the art can add one or more other operations to the flowchart or remove one or more operations from the flowchart under the guidance of the content of the present application.

[0024] In addition, the described embodiments are only a part of the embodiments of the present application, rather than all embodiments. The components of the embodiments of the present application described and illustrated in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but merely represents the selected embodiments of the present application. All other embodiments obtained by those skilled in the art based on the embodiments of the present application without creative efforts fall within the protection scope of the present application.

[0025] In order to enable those skilled in the art to use the content of the present application, the following implementation manners are given in combination with a specific application scenario, "underground parking lot ventilation control". For those skilled in the art, without departing from the spirit and scope of the present application, the general principles defined here can be applied to other embodiments and application scenarios.

[0026] The methods, devices, electronic devices, or computer-readable storage media in the embodiments of the present application can be applied to any scenario that requires underground parking lot ventilation control. The embodiments of the present application do not limit the specific application scenarios. Any solution that uses the underground parking lot fan prediction control method and system provided in the embodiments of the present application falls within the protection scope of the present application.

[0027] It should be noted that before the present application was proposed, the operation modes of traditional ventilation systems were usually limited by fixed schedules or simple sensing mechanisms, lacking the flexibility to achieve precise regulation of ventilation strategies. The deficiencies of such systems are that they cannot effectively avoid the occurrence of insufficient or excessive ventilation, thus having a negative impact on air quality and the comfort of drivers.

[0028] To address the above problems, the embodiments of the present application determine the predicted ventilation strategies corresponding to multiple preset future time periods in the underground parking lot based on the vehicle entry and exit data of the underground parking lot, the vehicle flow data in each area, and a dynamic prediction model; based on the temperature and humidity change data of the underground parking lot, the air quality change data in each area, and a comfort feedback model, optimize the ventilation strategy for the target area of the underground parking lot to generate the target ventilation strategies corresponding to multiple preset future time periods in the underground parking lot. By predicting the ventilation requirements in preset future time periods and real-time sensing environmental changes, the present application dynamically adjusts the ventilation strategy, which can reduce energy consumption while improving the comfort of drivers.

[0029] For the convenience of understanding the present application, the technical solutions provided by the present application are described in detail below in combination with specific embodiments.

[0030] Figure 1 This is a flowchart of a method for predicting and controlling a fan in an underground parking lot provided by an embodiment of the present application. AsFigure 1 As shown in Figure 1 , the underground parking lot fan prediction control method provided by the embodiments of the present application is applied to a ventilation optimization device, and includes the following steps:

[0031] S101: Obtain the vehicle entry and exit data of the underground parking lot, the vehicle flow data of each area, and the temperature and humidity change data of the underground parking lot and the air quality change data of each area collected in real time by the data collection device within a first preset historical time period.

[0032] In specific implementation, considering that the traffic flow and environmental quality of the underground parking lot are different at different time periods. For example, there are significant differences in traffic flow during holiday periods and working days, and there are also significant differences in traffic flow during the morning rush hour and night time during working days. Moreover, the environmental quality corresponding to different traffic flows is also different. Therefore, the embodiments of the present application conclude that the ventilation demand of the underground parking lot is highly dependent on time control and the surrounding environment.

[0033] In this regard, the embodiments of the present application capture the time change characteristics of the vehicle entry and exit data (such as vehicle entry and exit rates) and vehicle flow data of the underground parking lot within a historical period of time to predict the ventilation demand of the underground parking lot in a future period of time. In addition, based on the temperature and humidity change data of the underground parking lot and the air quality change data of each area, the ventilation effect and comfort level of each area of the underground parking lot are reflected, and corresponding feedback control strategies are obtained, so as to formulate target ventilation strategies corresponding to multiple preset future time periods.

[0034] Here, the embodiments of the present application deploy a variety of high-precision sensor networks for each area of the underground parking lot, that is, data collection devices, which can conveniently collect multi-dimensional information in the parking lot in real time, such as vehicle entry and exit data of the underground parking lot, temperature and humidity change data, vehicle flow data of each area, air quality change data, etc. This multi-dimensional scene perception ability can deeply understand the actual situation in the underground parking lot and can intelligently feedback and adjust the ventilation strategy according to real-time data. Compared with the single sensor or simple induction mechanism relied on by traditional systems, the embodiments of the present application have significant advantages in the breadth and depth of scene perception, not only can ensure a quick response to achieve precise control, but also can achieve energy conservation and emission reduction while ensuring the safety and comfort of drivers and pedestrians.

[0035] Specifically, vehicle entry and exit data of the underground parking lot can be collected based on vehicle entry and exit detectors, and vehicle flow data of each area can be collected based on infrared sensors deployed in each area of the underground parking lot. Thus, through the above data ventilation optimization device, the entry and exit rates of the vehicle flow in the entire underground parking lot and the vehicle flow conditions in each area of the underground parking lot can be analyzed, which can reflect the different usage intensities of each area inside the underground parking lot. In addition, air quality change data of each area of the underground parking lot can be collected based on CO concentration sensors deployed in each area of the underground parking lot, and temperature and humidity change data of the underground parking lot can be collected based on temperature and humidity sensors deployed in the central area of the underground parking lot.

[0036] S102: Based on the vehicle entry and exit data of the underground parking lot, the vehicle flow data of each area, and the dynamic prediction model, determine the predicted ventilation strategies corresponding to the underground parking lot in multiple preset future time periods; the predicted ventilation strategies include ventilation strategies for each area.

[0037] In a specific implementation, the embodiment of the present application captures the time change characteristics of the vehicle entry and exit data (such as vehicle entry and exit rates) and vehicle flow data of the underground parking lot within a historical period of time through a dynamic prediction model to predict the ventilation demand of the underground parking lot in a future period of time. Moreover, the data set can be updated in real time through the data collected by the vehicle entry and exit detectors and infrared sensors in the embodiment of the present application, which can ensure the accuracy of the predicted data.

[0038] Among them, the dynamic prediction model can be a neural network model, such as a Long Short-Term Memory Network (LSTM). The dynamic prediction model can specifically process time series data. The first preset historical period can be set according to actual requirements such as operation rate, response time, and control accuracy. Generally, the first preset historical period can be set to a time length of several months.

[0039] Exemplarily, by analyzing the vehicle entry and exit data of the entire underground parking lot every 5 minutes in the previous 6 months (the first preset historical period) and the vehicle flow data of each area in the underground parking lot in the previous 6 months, the dynamic prediction model can predict the vehicle entry and exit data of the entire underground parking lot and the vehicle flow data of each area in the future 15 minutes, 30 minutes, and 1 hour (multiple preset future time periods), so as to obtain the corresponding predicted ventilation strategies.

[0040] Here, the embodiments of the present application can effectively capture the dynamic changes of time series data by adopting an improved long short-term memory (LSTM) neural network model, which can improve the prediction accuracy. On this basis, it can accurately predict the ventilation requirements for multiple preset future time periods, such as multiple future 15-minute, 30-minute, and 1-hour periods, and accordingly turn on the fan circuit according to the requirements. This dynamic prediction mechanism completely gets rid of the static or rule-driven mode relied on by traditional ventilation systems, enabling the ventilation strategy to be adjusted prospectively according to future scenario changes, thereby achieving highly autonomous energy-saving prediction control.

[0041] S103: Based on the temperature and humidity change data of the underground parking lot, the air quality change data of each area, and the comfort feedback model, optimize the ventilation strategy for the target area of the underground parking lot, generate the target ventilation strategies corresponding to the underground parking lot in multiple preset future time periods respectively, and send the target ventilation strategies to the fan control device.

[0042] In specific implementation, the embodiments of the present application also collect the air quality data of each area of the underground parking lot through pre-deployed CO concentration sensors, and collect the temperature and humidity data of the underground parking lot through sensors such as temperature and humidity sensors, and transmit the collected data to the comfort feedback model to analyze the change trends of the air quality data (such as CO concentration) and the temperature and humidity data, so as to ensure that when the temperature and humidity data and the air quality data of the underground parking lot change greatly, the ventilation strategy can be adjusted dynamically and in a timely manner, and finally form a closed-loop control to maintain the best ventilation effect and energy-saving state. That is to say, the embodiments of the present application can receive the feedback information of multi-variable time series in real time (the temperature and humidity change data of the underground parking lot, the air quality change data of each area), and can use the identified air quality change and temperature and humidity change to judge the quality of the ventilation effect, that is, judge the impact of ventilation on air circulation and comfort, and jointly analyze the temperature and humidity, air quality and ventilation requirements, which is convenient for further improving ventilation under more complex environmental conditions in the future, and determining whether it is necessary to adjust the ventilation strategy to further optimize the strategy execution process. Specifically, the comfort feedback model mainly analyzes the data such as the CO concentration, temperature, and humidity collected by the CO concentration sensor and the temperature and humidity sensor per minute, so as to reflect the ventilation effect and comfort of each area of the underground parking lot, obtain the corresponding feedback control strategy, and then determine the target ventilation strategies corresponding to each preset future time period respectively.

[0043] Here, when designing the embodiments of the present application, full consideration is given to the different functionalities of each area, and data such as traffic flow and CO concentration also vary accordingly. By collecting data such as traffic flow, CO concentration, temperature, and humidity in each area and introducing them into the feedback closed-loop optimization of the system, the system can dynamically adjust and optimize the ventilation strategy again according to the feedback data of users to ensure meeting the requirements of different areas for the comfort of air circulation. Different from the unidirectional control mode of traditional ventilation systems, the embodiments of the present application adopt a closed-loop optimization strategy of parking lot data feedback, enabling the state of each area to be reflected in ventilation control in a timely manner. While meeting the energy-saving goal, it significantly improves the precise control of each area and creates a more comfortable parking experience.

[0044] It should be noted that most of the ventilation systems in traditional underground parking lots adopt fixed-time control, and a few parking lots adopt simple induction control, which cannot be dynamically adjusted according to the actual usage situation, resulting in a large amount of unnecessary energy consumption waste. Especially during periods with low traffic flow, the ventilation equipment still operates at high power. By introducing an intelligent prediction algorithm based on historical data and real-time data, the embodiments of the present application can predict future traffic flow and air quality requirements in advance and formulate ventilation strategies in advance, which can significantly reduce ineffective ventilation and lower energy consumption.

[0045] In addition, considering that traditional ventilation systems often cannot flexibly adjust the ventilation intensity according to real-time traffic flow and environmental changes, resulting in poor or excessive air circulation in some areas, affecting the breathing comfort and health safety of drivers. Through multi-sensor data fusion technology, the embodiments of the present application can monitor the air quality, temperature, humidity, and traffic dynamics of the parking lot in real time to ensure providing the best air circulation environment at different times and scenarios, improving the breathing comfort of drivers and reducing health risks.

[0046] In the embodiments of the present application, based on the vehicle entry and exit data of the underground parking lot, the vehicle flow data of each area, and the dynamic prediction model, the predicted ventilation strategies corresponding to multiple preset future time periods of the underground parking lot are determined; based on the temperature and humidity change data of the underground parking lot, the air quality change data of each area, and the comfort feedback model, the ventilation strategy of the target area of the underground parking lot is optimized to generate the target ventilation strategies corresponding to multiple preset future time periods of the underground parking lot. By predicting the ventilation requirements in preset future time periods and perceiving environmental changes in real time, the present application dynamically adjusts the ventilation strategy, which can reduce energy consumption while improving the comfort of drivers.

[0047] In a possible implementation manner, the predicted ventilation strategy includes one of a low-demand strategy, a medium-demand strategy, and a high-demand strategy; determining the predicted ventilation strategy corresponding to the underground parking lot in multiple preset future time periods based on the vehicle in-and-out data of the underground parking lot, the vehicle flow data of each area, and the dynamic prediction model in step S102 includes the following steps:

[0048] Step 1021: Extract time features and traffic flow features from the vehicle in-and-out data of the underground parking lot and the vehicle flow data of each area; the time features include weekday features and holiday features.

[0049] Step 1022: Input the traffic flow features and the time features into the dynamic prediction model to determine the vehicle in-and-out volume levels corresponding to the underground parking lot in multiple preset future time periods.

[0050] Step 1023a: If the vehicle in-and-out volume level corresponding to any of the preset future time periods is less than or equal to the first level, determine that the predicted ventilation strategy corresponding to the preset future time period is the low-demand strategy; the low-demand strategy includes opening the first number of ventilation ducts in each area and closing all the exhaust ducts in all areas.

[0051] Step 1023b: If the vehicle in-and-out volume level corresponding to any of the preset future time periods is greater than the first level and less than or equal to the second level, determine that the predicted ventilation strategy corresponding to the preset future time period is the medium-demand strategy; the medium-demand strategy includes opening the second number of ventilation ducts in each area and opening the exhaust ducts corresponding to the areas where the traffic flow is greater than the preset threshold; the first number is less than the second number.

[0052] Step 1023c: If the vehicle in-and-out volume level corresponding to any of the preset future time periods is greater than the second level and less than or equal to the third level, determine that the predicted ventilation strategy corresponding to the preset future time period is the high-demand strategy; the high-demand strategy includes opening all the ventilation ducts in each area, opening the exhaust ducts at the entrance of each area, and opening all the exhaust ducts corresponding to the areas where the traffic flow is greater than the preset threshold.

[0053] It should be noted that since the ventilation requirements of the underground parking lot vary at different times, and there are also differences in the ventilation requirements of each area of the underground parking lot during the same period, it can be concluded that the ventilation requirements of the underground parking lot highly depend on time control and the surrounding environment. In this regard, in the embodiments of the present application, for the time characteristics that highly depend on, the time characteristics are mainly extracted into two parts: working day characteristics and holiday characteristics. And the environmental characteristics directly affect the ventilation requirements of the parking lot. The system extracts the environmental characteristics of the underground parking lot using sensor data, mainly including traffic flow characteristics, air quality characteristics, and temperature and humidity characteristics. Statistical methods such as the Pearson correlation coefficient can be used to evaluate the correlation between each characteristic and the ventilation requirements, and the most valuable characteristics can be selected from them.

[0054] Among them, for the working day characteristics, the working day characteristics are extracted from the vehicle entry and exit data of the underground parking lot and the vehicle flow data of each area, and the time period characteristics such as the morning and evening rush hours and the night low valley from Monday to Friday are identified. These characteristics are extracted through the discrete Fourier transform technology, which can accurately describe the traffic flow and ventilation requirement fluctuations within a working day. For the holiday characteristics, through special analysis of the data, the system identifies the differences in ventilation requirements at special time points such as weekends and holidays. Capturing the holiday characteristics in the complex periodic changes is also extracted through the discrete Fourier transform technology to improve the accuracy of the prediction model. For the traffic flow characteristics, based on the data of vehicle entry and exit detectors and infrared sensors, the traffic flow entry and exit rate and vehicle distribution are analyzed to reflect the usage intensity of each area inside the underground parking lot, and different traffic flow patterns (no vehicle 0, very few 1, few 2, normal 3, many 4, very many 5) are identified through clustering data, and these patterns are used as the input of the dynamic prediction model.

[0055] In specific implementation, the target ventilation strategies formulated for the underground parking lot may include three schemes: low demand strategy, medium demand strategy, and high demand strategy.

[0056] Exemplarily, for the low demand strategy: when the vehicle entry and exit volume level of the entire underground parking lot in the next 15 minutes (a preset future time period) is predicted to be level 0 or 1 (mainly at night time) through the dynamic prediction model, it is determined that the predicted ventilation strategy corresponding to the next 15 minutes is the low demand strategy. Specifically, the low demand strategy can be to turn on 1 ventilation duct in the middle of each area in advance to reduce the ventilation intensity, and close the exhaust ducts of all areas, so as to save energy to the greatest extent while maintaining good air quality.

[0057] Exemplarily, for the medium demand strategy: when the dynamic prediction model can be used to predict that the vehicle inlet and outlet volume of the entire parking lot will be level 2 and 3 (mainly during daytime hours) in the next 15 minutes, the predicted ventilation strategy corresponding to the next 15 minutes is determined to be the medium demand strategy. Specifically, the medium demand strategy can be to open two ventilation ducts in the middle and near the entrance of each area in advance to increase the ventilation intensity, and when it is predicted that the traffic volume in a certain area (A, B, C, D areas) is large, especially for some areas with large traffic volume, the exhaust ducts at the entrance of the area will be opened in advance to achieve the purpose of air circulation in a certain area, so as to ensure that energy waste is avoided and energy is saved while maximizing the maintenance of good air quality.

[0058] Exemplarily, for the high-demand strategy: when the dynamic prediction model predicts that the vehicle inlet and outlet volume of the entire parking lot in the next 15 minutes will be level 4 and 5 (mainly during peak hours), the predicted ventilation strategy corresponding to the next 15 minutes is determined to be a high-demand strategy. Specifically, the high-demand strategy may be to open all ventilation ducts in each area to increase the ventilation intensity and open the exhaust ducts at the entrance of the area in advance. When it is predicted that the traffic volume in a certain area (A, B, C, D area) is the largest, especially for some areas with large traffic volume, all exhaust ducts will be opened in advance to ensure air circulation in a certain area and ensure that the air circulation environment in the parking lot is suitable.

[0059] For example, Figure 2 The schematic diagram of the overall area of ​​the underground parking lot and the sensor installation locations provided by the embodiment of the present application is shown. Figure 2 As shown, the underground parking lot includes four areas, A, B, C, and D. Vehicle entry and exit detectors are installed at the entrance and exit of the underground parking lot. Each area is installed with multiple infrared sensors for detecting vehicle flow data in the area.

[0060] For example, Figure 3 A schematic diagram showing any area and sensor installation location in an underground parking lot provided in an embodiment of the present application is shown; Figure 3 As shown, the ventilation equipment in this area mainly includes ventilation fans, exhaust fans and intelligent controllers. Specifically, this area contains three ventilation ducts and two exhaust ducts. Multiple ventilation fans are arranged on each ventilation duct, and multiple exhaust fans are arranged on each exhaust duct. These ventilation fans and exhaust fans are evenly distributed in various locations in various areas of the parking lot to ensure that all areas can be covered.

[0061] In a possible implementation manner, the dynamic prediction model is a long short-term memory network model; the dynamic prediction model includes an input layer, an LSTM layer, a fully connected layer, and an output layer; the step of inputting the traffic flow feature and the time feature into the dynamic prediction model in step 1022 to determine the vehicle in-and-out volume levels corresponding to the underground parking lot in a plurality of preset future time periods includes the following steps:

[0062] Input the traffic flow feature and the time feature into the input layer to determine the time feature and the traffic flow feature with the same feature dimension; input the time feature and the traffic flow feature with the same feature dimension into the LSTM layer to determine the intermediate dependent feature; input the intermediate dependent feature into the fully connected layer to determine the vehicle in-and-out volume corresponding to the underground parking lot in a plurality of preset future time periods and the predicted values of the traffic flow of each area; input each predicted value into the output layer to determine the vehicle in-and-out volume levels corresponding to the underground parking lot in a plurality of preset future time periods.

[0063] In a specific implementation, the ventilation optimization device provided in the embodiment of the present application internally uses a long short-term memory network model as the core dynamic prediction model. The LSTM network model can effectively capture the long-term dependence relationship in the pre-organized time series data, and is very suitable for processing the time series data of the underground parking lot. The dynamic prediction model structure mainly includes multiple neural networks such as an input layer, an LSTM layer, a fully connected layer, and an output layer. The ventilation optimization device trains the model by running the dynamic prediction model and inputting a large amount of historical data, and the goal is to minimize the error between the predicted value and the actual value. The model is updated as the usage situation of the underground parking lot changes, mainly by re-training the LSTM model by updating the regular data to the latest data. The focus of the embodiment of the present application is to use continuously updated data and training to enable the model to gradually adapt to the new data features on the basis of not losing historical data.

[0064] It should be noted that for the input layer, the dynamic prediction model receives the preprocessed time features and environmental features. By inputting the data of various sensors and performing normalization processing, it ensures that each feature is in the same magnitude for model training, which helps to improve the convergence speed and prediction accuracy of the model. For the LSTM layer, multiple LSTM units are designed, and the connection method between layers can effectively memorize the dependencies in the long time series. By introducing bidirectional LSTM, the context information of the time series can be considered simultaneously, improving the prediction accuracy. For the fully connected layer, the output of the LSTM layer is mapped to one or more predicted values through the fully connected layer. This layer uses linear or non-linear activation functions to enhance the expression ability of the model. For the output layer, according to the specific application scenario, the predicted values of the vehicle inflow and outflow in the next 15 minutes, 30 minutes, and 1 hour, as well as the traffic flow in each area, are finally output, and the corresponding ventilation requirements are provided to the fan.

[0065] Here, the input layer can adopt Min-Max normalization and Z-score standardization techniques. Among them, the Min-Max normalization technique is a technique that scales the data proportionally to a specific range (usually [0, 1] or [-1, 1]). By this method, the minimum value in the original data is mapped to the minimum value of the specified range, the maximum value is mapped to the maximum value of the specified range, and other values are linearly scaled proportionally. The calculation formula of Min-Max normalization is as follows: ; where, is the original vehicle inflow and outflow data, is the normalized data, and are the minimum and maximum values of the original data respectively, and are the minimum and maximum values of the target range.

[0066] In addition, the Z-score standardization technique is a method that converts the data into a standard normal distribution with a mean of 0 and a standard deviation of 1. By this method, each value in the data is expressed as the multiple of the deviation of it from the mean relative to the standard deviation, thus eliminating the influence of the dimension of the original data. The calculation formula of Z-score standardization is as follows: ; where, is the original vehicle inflow and outflow data, is the standardized data value, is the mean of the original data, is the standard deviation of the original data.

[0067] It should also be noted that the Pearson correlation coefficient is a statistical indicator used to measure the strength and direction of the linear relationship between two variables. Its value range is from -1 to 1, indicating perfect negative correlation (-1), no correlation (0), and perfect positive correlation (1). When evaluating the correlation between each feature and the ventilation demand, the Pearson correlation coefficient can help select those features that are most valuable for predicting the ventilation demand. Suppose there are two variables and , representing a certain feature and the ventilation demand respectively. The calculation formula of the Pearson correlation coefficient is as follows: ; ; where, is the variable, and are the variables, and represent the th observation values respectively, and are the means of and respectively, is and the Pearson correlation coefficient between them, is the number of observation values.

[0068] In addition, when analyzing the fluctuations of traffic flow and ventilation demand during weekdays, the Fourier transform can be used to extract and describe the frequency components of these fluctuations. The time series data represents the changes in traffic flow or ventilation demand within a certain time range (such as weekdays within a week). The traffic flow data is usually discrete, that is, there are M sampling points , where, T is the sampling period, m = 0, 1, 2,..., M - 1. The calculation formula of the discrete Fourier transform is as follows: ; where, is the th frequency component representing the component with frequency in the signal, is the discrete number of vehicles in the time domain, and M is the number of sampling points of the traffic flow.

[0069] It should be noted that the dynamic prediction model adopted in the embodiments of this application is an improved LSTM network model, and the calculation formula of the dynamic prediction model is:

[0070] ; ; where, represents at time step The input feature vector represents the normalized traffic flow data and other relevant features collected by sensors. Denoted as the hidden state at the previous time step which represents the information of the previous moment memorized by the model. Denoted as the hidden state at the previous time step The cell state stores the long-term memory of the previous moment. Denoted as at time step The output of the forget gate, representing the proportion of information in the cell state that needs to be "forgotten", with its value ranging from [0, 1], where 1 means complete retention and 0 means complete forgetting. Denoted as at time step The output of the input gate, which determines how much new information at the current time step needs to be added to the cell state, with its value ranging from [0, 1]. Denoted as the candidate cell state, which is the candidate value of the new information generated by the tanh activation function, with its value ranging from [-1, 1]. Denoted as at time step The cell state at the current time step is the long-term memory at the current time step, which is jointly determined by the cell state at the previous time step and the current input gate, forget gate, and candidate cell state. Denoted as at time step The output of the output gate, which determines how much information is extracted from the cell state as the hidden state output at the current time step, with its value ranging from [0, 1]. Denoted as the hidden state at the time step (also the output of the LSTM cell), which combines the current cell state and the output of the output gate and is mapped to the range [-1, 1] through the tanh activation function. , , , are the weight matrices of the forget gate, input gate, candidate cell state, and output gate respectively, which transform the hidden state at the previous time step and the input at the current time step into the outputs of the corresponding gates. , , , are the bias vectors of the forget gate, input gate, candidate cell state, and output gate respectively, which act in the calculation of the corresponding gate outputs. Denoted as the sigmoid activation function, defined as , and its output range is [0, 1], which is often used to calculate the output values of the gates. Tanh is denoted as the hyperbolic tangent activation function, defined as , whose output range is between [-1, 1], and is often used to process cell states and hidden states. Indicates at time step The final predicted value, representing the vehicle in-and-out volume or ventilation demand in the next 15 minutes, 30 minutes, and 1 hour. The weight matrix of the fully connected layer, which maps the output of the LSTM layer to the predicted value. Is represented as the bias vector of the fully connected layer.

[0071] In a possible implementation, the dynamic prediction model is trained according to the following steps: Obtain a plurality of historical sample data and the actual value corresponding to each historical sample data; the historical sample data includes the vehicle in-and-out data of the underground parking lot collected within a second preset future time period, and the vehicle flow data of each area; the actual value is the actual vehicle in-and-out volume level corresponding to a preset actual time period; input each historical sample data into an initial long short-term memory network model to determine the predicted value corresponding to each historical sample data; the predicted value is the predicted vehicle in-and-out volume level corresponding to a preset actual time period; based on the loss between the predicted value and the actual value of each historical sample data, update the network parameters of the initial long short-term memory network model to obtain the dynamic prediction model.

[0072] Here, the backpropagation algorithm can be used to adjust the internal parameters of the model, and the optimization goal is to reduce the mean square error loss function. Specifically, grid search and Bayesian optimization automatic hyperparameter tuning techniques can also be used to optimize the hyperparameters of the model, such as the number of layers, the number of units, the learning rate, etc., to ensure the best performance of the model.

[0073] In a possible implementation, the air quality change data includes the average CO concentration change data; in step S103, based on the temperature and humidity change data of the underground parking lot, the air quality change data of each area, and the comfort feedback model, optimizing the ventilation strategy of the target area of the underground parking lot to generate the target ventilation strategies corresponding to the underground parking lot in multiple preset future time periods, including the following situations:

[0074] Situation 1: If the rising rate of the average CO concentration in the target area is greater than or equal to the first rate threshold, add one more ventilation duct opened in the target area, and when it is detected that all the ventilation ducts in the target area have been opened, additionally open one exhaust duct.

[0075] Situation 2: If the falling rate of the average CO concentration in the target area is less than or equal to the second rate threshold, close all the exhaust ducts in the target area, and when it is detected that all the exhaust ducts in the target area have been closed, additionally close one ventilation duct.

[0076] In a specific implementation, the embodiment of the present application provides feedback-regulated ventilation control. On the premise of ensuring that the entire underground parking lot area can normally predict and activate the above three predicted ventilation strategies in advance, the ventilation strategy of a single area is further adjusted appropriately according to the magnitude of the CO concentration change data collected by the CO concentration sensor, mainly including the following situations. When the average CO concentration in a certain area rises rapidly, the number of opened ventilation ducts in this area can be increased. When the average CO concentration in a certain area drops rapidly, the number of opened ventilation ducts in this area can be reduced. Ensure that each CO concentration will be maintained in a stable range, and require its overshoot to be as low as possible to achieve the most comfortable environment for the human body.

[0077] Specifically, when the average CO concentration in a certain area rises rapidly, when it is detected that the CO concentration rising rate in a certain area is greater than or equal to the first speed threshold, such as 20 PPM / min, one more ventilation duct in this area will be opened. When all the ventilation ducts are already opened, one more exhaust duct will be opened to ensure comfort while saving energy to the greatest extent.

[0078] Specifically, when the average CO concentration in a certain area drops rapidly, when it is detected that the CO concentration dropping rate in a certain area is less than or equal to the second rate threshold, such as 20 PPM / min, all the exhaust ducts in this area will be closed. When all the exhaust ducts in this area are already closed, one ventilation duct will be closed to ensure comfort while saving energy to the greatest extent.

[0079] It should be noted that the traditional ventilation control method lacks an intelligent management method and is usually manually controlled by operators, unable to deeply analyze and predict the actual usage situation of the parking lot in a timely manner, resulting in control lag and low efficiency. However, the embodiment of the present application can realize the real-time analysis and prediction of the parking lot usage situation through the neural network prediction algorithm, provide a dynamically adjustable ventilation strategy, control the operation state and wind force of the fans in each area, and achieve the best balance between energy saving and comfort. Specifically, it can accurately sense and predict the actual traffic flow and air quality changes, formulate a reasonable target ventilation strategy, avoid the situation of insufficient or excessive ventilation, achieve an ideal energy-saving effect, and ensure the air circulation and comfort in the underground parking lot.

[0080] Based on the same inventive concept, the embodiment of the present application also provides an underground parking lot fan prediction control corresponding to the underground parking lot fan prediction control method provided in the above embodiment. Since the principle of solving problems by the system in the embodiment of the present application is similar to that of the underground parking lot fan prediction control method in the above embodiment of the present application, the implementation of the system can refer to the implementation of the method, and the repeated parts will not be elaborated.

[0081] Figure 4 shows a schematic structural diagram of the underground parking lot fan prediction control system 400 provided by an embodiment of the present application. As Figure 4 shown, the underground parking lot fan prediction control system 400 includes a data acquisition device 410, a ventilation optimization device 420, and a fan control device 430. The ventilation optimization device 420 is communicatively connected to the data acquisition device 410 and the fan control device 430 respectively; wherein,

[0082] the data acquisition device 410 is configured to collect the vehicle in-and-out data, temperature and humidity change data at each moment, as well as the vehicle flow data and air quality change data in each area of the underground parking lot;

[0083] the ventilation optimization device 420 is configured to determine the predicted ventilation strategies corresponding to the underground parking lot in a plurality of preset future time periods based on the vehicle in-and-out data, vehicle flow data in each area, and a dynamic prediction model of the underground parking lot in a first preset historical time period. The predicted ventilation strategies include the ventilation strategies for each area; based on the real-time temperature and humidity change data, air quality change data in each area, and a comfort feedback model of the underground parking lot, optimize the ventilation strategy of the target area of the underground parking lot, generate the target ventilation strategies corresponding to the underground parking lot in a plurality of preset future time periods, and send the target ventilation strategies to the fan control device 430;

[0084] the fan control device 430 is configured to control the ventilation equipment in each area of the underground parking lot to ventilate according to the ventilation control instructions corresponding to the target ventilation strategy.

[0085] It should be noted that the embodiment of the present application relates to an underground parking lot fan prediction control system 400, which realizes efficient management of parking lot ventilation equipment through the integrated operation of multi-dimensional data, intelligent prediction, and real-time control. The underground parking lot fan prediction control system 400 includes a data acquisition device 410, a ventilation optimization device 420, and a fan control device 430. The data acquisition device 410 serves as the perception layer of the intelligent ventilation control system, responsible for collecting multi-dimensional environmental data in the parking lot in real time and performing preliminary processing through a multi-dimensional scenario perception model. This system mainly consists of an infrared sensor module, a CO concentration sensor module, a temperature and humidity sensor module, a vehicle detector module, and a data processing module. The ventilation optimization device 420 is the core part of the intelligent ventilation control, mainly including a real-time dynamic prediction system and a comfort feedback model, responsible for analyzing and predicting the collected multi-dimensional data and generating specific ventilation strategies. The fan control device 430 is the execution layer of the system, including a control module and ventilation equipment, responsible for applying the optimized ventilation strategies to the actual ventilation equipment. Generally speaking, the system is based on the real-time collected multi-dimensional environmental data, generates ventilation strategies by fusing time series prediction and dynamic generation. The real-time response module executes these strategies and optimizes them according to user feedback and environmental changes. The entire system forms a closed-loop control mechanism, continuously improving the energy-saving effect and user comfort of the system.

[0086] In a possible implementation manner, the predicted ventilation strategy includes one of a low-demand strategy, a medium-demand strategy, and a high-demand strategy; specifically, the ventilation optimization device 420 is configured to determine the predicted ventilation strategies corresponding to the underground parking lot in multiple preset future time periods according to the following steps: extract time features and traffic flow features from the vehicle entry and exit data of the underground parking lot and the vehicle flow data of each area; the time features include weekday features and holiday features; input the traffic flow features and the time features into the dynamic prediction model to determine the vehicle entry and exit volume levels corresponding to the underground parking lot in multiple preset future time periods respectively; if the vehicle entry and exit volume level corresponding to any one of the preset future time periods is less than or equal to the first level, determine that the predicted ventilation strategy corresponding to the preset future time period is the low-demand strategy; the low-demand strategy includes opening the first number of ventilation ducts in each area and closing the exhaust ducts of all areas; if the vehicle entry and exit volume level corresponding to any one of the preset future time periods is greater than the first level and less than or equal to the second level, determine that the predicted ventilation strategy corresponding to the preset future time period is the medium-demand strategy; the medium-demand strategy includes opening the second number of ventilation ducts in each area and opening the exhaust ducts corresponding to the areas where the traffic flow is greater than the preset threshold; the first number is less than the second number; if the vehicle entry and exit volume level corresponding to any one of the preset future time periods is greater than the second level and less than or equal to the third level, determine that the predicted ventilation strategy corresponding to the preset future time period is the high-demand strategy; the high-demand strategy includes opening all the ventilation ducts in each area, opening the exhaust ducts at the entrance of each area, and opening all the exhaust ducts corresponding to the areas where the traffic flow is greater than the preset threshold.

[0087] In a possible implementation manner, the data acquisition device 410 includes an infrared sensor, a CO concentration sensor, a temperature and humidity sensor, a vehicle entry and exit detector, and a data processing module; different vehicle entry and exit detectors are respectively installed at the entrance and exit of the underground parking lot, and any one of the vehicle entry and exit detectors includes a vehicle identification camera and a geomagnetic inductor, the vehicle identification camera is used to identify the vehicles entering and leaving the parking lot, and the geomagnetic inductor is used to sense the entry and exit of the vehicles; different infrared sensors are installed at the connection points of the driving lanes in each area of the underground parking lot; different CO concentration sensors are evenly distributed in each area of the underground parking lot; the temperature and humidity sensor is deployed in the central area of the underground parking lot; the data processing module includes arranging a low-pass filter and a median filter on the vehicle entry and exit detector, and arranging a Kalman filter on the infrared sensor; wherein,

[0088] the vehicle entry and exit detector is used to obtain the vehicle entry and exit data of the underground parking lot in real time;

[0089] The infrared sensor is used to detect the dynamic activity data of vehicles and pedestrians in each area; the dynamic activity data includes vehicle flow data and pedestrian flow data;

[0090] The CO concentration sensor is used to collect the corresponding average CO concentration in different areas in real time;

[0091] The temperature and humidity sensor is used to collect the temperature and humidity change data of the entire underground parking lot in real time;

[0092] The data processing module is used to filter the random noise data in the target data collected by each sensor, and send the filtered target data to the ventilation optimization device.

[0093] Here, vehicle entry and exit detectors are mainly arranged at each entrance and exit of the underground parking lot, and each vehicle entry and exit detector is communicatively connected to the ventilation optimization device to ensure that these devices can monitor the entry and exit of vehicles in real time, provide accurate and stable vehicle flow data (providing vehicle entry and exit data every 5 minutes), and provide accurate data support for predicting ventilation requirements. Combining big data analysis and machine learning algorithms, the historical vehicle flow data is deeply mined to predict the future vehicle flow trend. The infrared sensor is mainly used to identify and classify different types of moving objects (pedestrians and cars can be distinguished), so as to accurately judge the vehicle flow and pedestrian flow in each area of the parking lot, improving the accuracy of vehicle flow and pedestrian flow monitoring. The CO concentration sensor provides data to help the fan system improve the ventilation strategy by monitoring the CO concentration in real time. Among them, the air detection in multiple dimensions is also distinguished, and the average CO concentration corresponding to different areas can be distinguished, so as to feedback to the control system to provide more detailed ventilation control. The temperature and humidity sensor can not only help evaluate the impact of the ventilation of the fan on the air quality, but also combine the data of other sensors to provide an important reference for the formulation of the ventilation strategy. The introduction of the temperature and humidity sensor is mainly to ensure that the system can adjust the ventilation strategy accordingly under different seasons and humid weather conditions, so as to provide a comfortable air circulation environment.

[0094] It should also be noted that the embodiment of the present application further includes a data preprocessing stage, which ensures the accuracy and consistency of data by introducing a variety of devices and technologies. Low-pass filters, median filters and Kalman filters are arranged on the sensors, which can effectively filter the random noise data collected by the sensors. At the same time, statistical analysis methods need to be adopted in the system to automatically identify and process extreme values caused by faults or data loss in the sensors, ensuring the stability and reliability of data input. At the same time, in order to avoid the influence of the difference in the magnitude of data characteristics of different sensors on model training, all feature values of the system are converted to the same magnitude, which helps to improve the convergence speed and prediction accuracy of the model.

[0095] In a possible implementation manner, the fan control device 430 includes a receiving control module and multiple ventilation devices; the receiving control module is communicatively connected to the ventilation devices in each area respectively; the ventilation devices include ventilation fans, exhaust fans and intelligent regulators; wherein,

[0096] The receiving control module is configured to determine the number of ventilation fans and exhaust fans to be turned on in each area according to the ventilation control instruction sent by the ventilation optimization device, and send the number of ventilation fans and exhaust fans to be turned on in each area to the intelligent regulator in the corresponding area;

[0097] The intelligent regulator is configured to control the corresponding number of ventilation fans and exhaust fans in the area to be turned on according to the number of ventilation fans and exhaust fans to be turned on in the corresponding area.

[0098] In a specific implementation, the receiving control module is connected to all the ventilation devices in each area through communication wires and is connected to the intelligent regulator wirelessly, which also facilitates the management personnel to manually adjust the number of fans in real time, ensuring that the system can quickly respond to the external environment. After receiving the signal, the system immediately executes the adjustment of the ventilation devices in the parking lot. Real-time adjustment is performed according to the set strategy to ensure the energy conservation and comfort of the parking lot. At the same time, the control module can also receive feedback information in real time through the linkage with the comfort feedback model, and can further optimize the strategy execution process.

[0099] Specifically, the receiving control module is connected to the ventilation optimization device through a wireless control network, and receives the optimization strategy instruction for fan control. The receiving control module is connected to the ventilation devices in the underground parking lot by wire, including the number of ventilation fans and the number of exhaust fans to be turned on in each area. Thus, the fan state and the number of fans can be adjusted to ensure that the system can quickly respond to the external environment. The ventilation devices include ventilation fans, exhaust fans and intelligent regulators distributed in each area of the parking lot. These devices are remotely managed through the control module to ensure that the system can quickly respond to environmental changes and user needs.

[0100] Exemplarily, Figure 5 The working process schematic diagram of the fan prediction control system for an underground parking lot provided by an embodiment of the present application is shown. Specifically, first, multi-source data is collected, preprocessed, and feature extracted by each sensor included in the data collection device, and time features and environmental features are extracted, and a target ventilation strategy for controlling the action execution of the ventilation device is obtained through a dynamic prediction model and a comfort feedback model.

[0101] In the embodiments of the present application, an efficient prediction control system for an underground parking lot fan is constructed mainly through highly integrated data acquisition, intelligent prediction, and feedback optimization control. The working principle of the system is based on the real-time acquisition and processing of multi-dimensional data. Intelligent ventilation strategies are generated through time series prediction and multi-objective optimization algorithms, and these strategies are executed by a control module. The closed-loop feedback mechanism ensures the continuous optimization of the system, meeting the dual requirements of energy conservation and user comfort. The structural characteristics and connection relationships of each module ensure the efficiency and reliability of the system, enabling it to cope with the complex and changeable parking lot environment.

[0102] Based on the same inventive concept, the embodiments of the present application also provide a computer-readable storage medium. A computer program is stored on the computer-readable storage medium, and when the computer program is run by a processor, it executes the steps of the underground parking lot fan prediction control method provided in the above embodiments.

[0103] Specifically, the storage medium can be a general storage medium, such as a mobile disk, a hard disk, etc. When the computer program on the storage medium is run, it can execute the above underground parking lot fan prediction control method. By predicting the ventilation demand in a preset future time period and real-time sensing environmental changes, the ventilation strategy can be dynamically adjusted, which can reduce energy consumption while improving the comfort of the driver.

[0104] In the embodiments of the present application, when the computer program is run by a processor, it can also execute other machine-readable instructions to execute other methods described in the embodiments. For the specific method steps and principles of execution, refer to the description of the embodiments, and details will not be elaborated here.

[0105] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For another example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some communication interfaces. The indirect coupling or communication connection of the devices or units can be in an electrical, mechanical, or other form.

[0106] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or they can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0107] In addition, each functional unit in the embodiments provided in this application may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit.

[0108] If the above-mentioned function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs.

[0109] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. In addition, the terms "first", "second", "third", etc. are only used for descriptive distinction and cannot be understood as indicating or implying relative importance.

[0110] Finally, it should be noted that the above-mentioned embodiments are only specific implementation manners of this application, used to illustrate the technical solutions of this application, and are not intended to limit it. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: any person skilled in the art within the technical scope disclosed in this application can still modify the technical solutions described in the foregoing embodiments, or can easily think of changes, or make equivalent replacements for some of the technical features; and these modifications, changes, or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of this application. All should be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

1. A method for predictive control of a fan in an underground parking lot, characterized in that, Applied to a ventilation optimization device, the method includes: Obtaining the vehicle in-and-out data of the underground parking lot, the vehicle flow data of each area collected by the data acquisition device within the first preset historical time period, and the temperature and humidity change data of the underground parking lot and the air quality change data of each area collected in real time; Based on the vehicle in-and-out data of the underground parking lot, the vehicle flow data of each area, and the dynamic prediction model, determining the predicted ventilation strategies corresponding to the underground parking lot in multiple preset future time periods; the predicted ventilation strategies include the ventilation strategies for each area; The predicted ventilation strategy includes one of a low-demand strategy, a medium-demand strategy, and a high-demand strategy; The low-demand strategy includes opening the first number of ventilation ducts in each area and closing all the exhaust ducts in all areas; The medium-demand strategy includes opening the second number of ventilation ducts in each area, opening the exhaust ducts corresponding to the areas where the vehicle flow is greater than the preset threshold, and the positions of the opened exhaust ducts are the middle and the positions near the entrance of each area; the first number is less than the second number; The high-demand strategy includes opening all the ventilation ducts in each area, opening the exhaust ducts at the entrance of each area, and opening all the exhaust ducts corresponding to the areas where the vehicle flow is greater than the preset threshold; Based on the temperature and humidity change data of the underground parking lot, the air quality change data of each area, and the comfort feedback model, optimizing the ventilation strategy of the target area of the underground parking lot, generating the target ventilation strategies corresponding to the underground parking lot in multiple preset future time periods, and sending the target ventilation strategies to the fan control device; The air quality change data includes the average CO concentration change data; the optimizing the ventilation strategy of the target area of the underground parking lot based on the temperature and humidity change data of the underground parking lot, the air quality change data of each area, and the comfort feedback model, generating the target ventilation strategies corresponding to the underground parking lot in multiple preset future time periods, includes: If the rising rate of the average CO concentration in the target area is greater than or equal to the first rate threshold, increasing the number of opened ventilation ducts in the target area by one, and when it is detected that all the ventilation ducts in the target area have been opened, additionally opening one exhaust duct; If the falling rate of the average CO concentration in the target area is less than or equal to the second rate threshold, closing all the exhaust ducts in the target area, and when it is detected that all the exhaust ducts in the target area have been closed, additionally closing one ventilation duct.

2. The method according to claim 1, wherein The determining the predicted ventilation strategies corresponding to the underground parking lot in multiple preset future time periods based on the vehicle in-and-out data of the underground parking lot, the vehicle flow data of each area, and the dynamic prediction model, includes: Extracting time features and vehicle flow features from the vehicle in-and-out data of the underground parking lot and the vehicle flow data of each area; the time features include weekday features and holiday features; Input the traffic flow characteristics and the time characteristics into the dynamic prediction model to determine the vehicle in-and-out volume levels corresponding to the underground parking lot in multiple preset future time periods; If the vehicle in-and-out volume level corresponding to any one of the preset future time periods is less than or equal to the first level, determine that the predicted ventilation strategy corresponding to the preset future time period is a low-demand strategy; If the vehicle in-and-out volume level corresponding to any one of the preset future time periods is greater than the first level and less than or equal to the second level, determine that the predicted ventilation strategy corresponding to the preset future time period is a medium-demand strategy; If the vehicle in-and-out volume level corresponding to any one of the preset future time periods is greater than the second level and less than or equal to the third level, determine that the predicted ventilation strategy corresponding to the preset future time period is a high-demand strategy.

3. The method according to claim 2, characterized in that, The dynamic prediction model is a long short-term memory network model; the dynamic prediction model includes an input layer, an LSTM layer, a fully connected layer, and an output layer; the step of inputting the traffic flow characteristics and the time characteristics into the dynamic prediction model to determine the vehicle in-and-out volume levels corresponding to the underground parking lot in multiple preset future time periods includes: Input the traffic flow characteristics and the time characteristics into the input layer to determine the time characteristics and the traffic flow characteristics with the same feature dimension; Input the time characteristics and the traffic flow characteristics with the same feature dimension into the LSTM layer to determine the intermediate dependent characteristics; Input the intermediate dependent characteristics into the fully connected layer to determine the vehicle in-and-out volumes corresponding to the underground parking lot in multiple preset future time periods respectively, and the predicted values of the traffic flow in each area; Input each predicted value into the output layer to determine the vehicle in-and-out volume levels corresponding to the underground parking lot in multiple preset future time periods respectively.

4. The method according to claim 3, wherein The dynamic prediction model is trained according to the following steps: Obtain multiple historical sample data and the actual value corresponding to each historical sample data; the historical sample data includes the underground parking lot vehicle in-and-out data collected within the second preset future time period, and the vehicle flow data in each area; The actual value is the actual vehicle in-and-out volume level corresponding to the preset actual time period; Input each historical sample data into the initial long short-term memory network model to determine the predicted value corresponding to each historical sample data; the predicted value is the predicted vehicle in-and-out volume level corresponding to the preset actual time period; Update the network parameters of the initial long short-term memory network model based on the loss between the predicted values and the actual values of each historical sample data to obtain the dynamic prediction model.

5. An underground parking lot fan prediction control system, characterized in that The system includes a data collection device, a ventilation optimization device, and a fan control device, and the ventilation optimization device is communicatively connected to the data collection device and the fan control device respectively; wherein, The data collection device is used to collect the underground parking lot vehicle in-and-out data, temperature and humidity change data, vehicle flow data in each area, and air quality change data at each moment; The ventilation optimization device is used to determine the predicted ventilation strategies corresponding to multiple preset future time periods of the underground parking lot based on the vehicle entry and exit data, vehicle flow data in each area, and a dynamic prediction model of the underground parking lot. The predicted ventilation strategies include ventilation strategies for each area; The predicted ventilation strategy includes one of a low-demand strategy, a medium-demand strategy, and a high-demand strategy; The low-demand strategy includes opening the first number of ventilation ducts in each area and closing the exhaust ducts in all areas; The medium-demand strategy includes opening the second number of ventilation ducts in each area, opening the exhaust ducts corresponding to the areas where the vehicle flow is greater than a preset threshold, and the positions of the opened exhaust ducts are in the middle and near the entrance of each area; the first number is less than the second number; The high-demand strategy includes opening all the ventilation ducts in each area, opening the exhaust ducts at the entrance of each area, and opening all the exhaust ducts corresponding to the areas where the vehicle flow is greater than a preset threshold; Based on the real-time temperature and humidity change data of the underground parking lot, the air quality change data in each area, and a comfort feedback model, optimize the ventilation strategy of the target area of the underground parking lot, generate the target ventilation strategies corresponding to multiple preset future time periods of the underground parking lot, and send the target ventilation strategies to the fan control device; The optimizing the ventilation strategy of the target area of the underground parking lot based on the real-time temperature and humidity change data of the underground parking lot, the air quality change data in each area, and a comfort feedback model, and generating the target ventilation strategies corresponding to multiple preset future time periods of the underground parking lot includes: If the rising rate of the average CO concentration in the target area is greater than or equal to the first rate threshold, increase the number of opened ventilation ducts in the target area by one, and when it is detected that all the ventilation ducts in the target area have been opened, additionally open one exhaust duct; If the falling rate of the average CO concentration in the target area is less than or equal to the second rate threshold, close all the exhaust ducts in the target area, and when it is detected that all the exhaust ducts in the target area have been closed, additionally close one ventilation duct; The fan control device is used to control the ventilation equipment in each area of the underground parking lot to ventilate according to the ventilation control instructions corresponding to the target ventilation strategy.

6. The system according to claim 5, wherein The ventilation optimization device is specifically used to determine the predicted ventilation strategies corresponding to multiple preset future time periods of the underground parking lot according to the following steps: Extract time features and vehicle flow features from the vehicle entry and exit data, vehicle flow data in each area of the underground parking lot; the time features include weekday features and holiday features; Input the vehicle flow features and the time features into the dynamic prediction model to determine the vehicle entry and exit volume levels corresponding to multiple preset future time periods of the underground parking lot; If the vehicle in-out volume level corresponding to any of the preset future time periods is less than or equal to the first level, determine that the predicted ventilation strategy corresponding to the preset future time period is a low-demand strategy; If the vehicle in-out volume level corresponding to any of the preset future time periods is greater than the first level and less than or equal to the second level, determine that the predicted ventilation strategy corresponding to the preset future time period is a medium-demand strategy; If the vehicle in-out volume level corresponding to any of the preset future time periods is greater than the second level and less than or equal to the third level, determine that the predicted ventilation strategy corresponding to the preset future time period is a high-demand strategy.

7. The system according to claim 5, characterized in that, The data acquisition device includes an infrared sensor, a CO concentration sensor, a temperature and humidity sensor, a vehicle in-out detector, and a data processing module; different vehicle in-out detectors are respectively installed at the entrance and exit of the underground parking lot. Any one of the vehicle in-out detectors includes a vehicle identification camera and a geomagnetic inductor. The vehicle identification camera is used to identify the vehicles entering and leaving the parking lot, and the geomagnetic inductor is used to sense the entry and exit of vehicles; different infrared sensors are installed at the connection of the driving lanes in each area of the underground parking lot; different CO concentration sensors are evenly distributed in each area of the underground parking lot; the temperature and humidity sensor is deployed in the central area of the underground parking lot; the data processing module includes arranging a low-pass filter and a median filter on the vehicle in-out detector, and arranging a Kalman filter on the infrared sensor; where, The vehicle in-out detector is used to obtain the vehicle in-out data of the underground parking lot in real time; The infrared sensor is used to detect the dynamic activity data of vehicles and pedestrians in each area; the dynamic activity data includes vehicle flow data and human flow data; The CO concentration sensor is used to collect the average CO concentration corresponding to different areas in real time; The temperature and humidity sensor is used to collect the temperature and humidity change data of the entire underground parking lot in real time; The data processing module is used to filter the random noise data in the target data collected by each sensor, and send the filtered target data to the ventilation optimization device.

8. The system according to claim 5, wherein The fan control device includes a receiving control module and multiple ventilation devices; the receiving control module is respectively communicatively connected to the ventilation devices in each area; the ventilation devices include ventilation fans, exhaust fans, and intelligent regulators; where, The receiving control module is used to determine the number of ventilation fans and exhaust fans to be turned on in each area according to the ventilation control instruction sent by the ventilation optimization device, and send the number of ventilation fans and exhaust fans to be turned on in each area to the intelligent regulator in the corresponding area; The intelligent regulator is used to control the corresponding number of ventilation fans and exhaust fans in the area to be turned on according to the number of ventilation fans and exhaust fans to be turned on in the corresponding area.

9. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, and when the computer program is run by a processor, it executes the steps of the underground parking lot fan prediction control method according to any one of claims 1 to 4.

Images