Metro station water chilling unit energy-saving control method and system

Abstract

The invention provides an energy-saving control method and system for a water chilling unit of a subway station, and particularly relates to the technical field of air conditioning systems.The method comprises the steps that passenger flow data, indoor and outdoor environment data and equipment data are obtained, load prediction is conducted, and equipment operation parameters are generated according to a load prediction result; the passenger flow data comprises personnel retention data; in a preset time period, forecast temperature data are obtained, whether at least one water chilling unit is shut down or not is judged according to the forecast temperature data and the personnel retention data, and if the water chilling unit is shut down, the phase change cold storage system is adopted for auxiliary cold supply; and regenerating equipment operation parameters according to the equipment data after auxiliary cooling, the passenger flow data and the indoor and outdoor environment data. According to the method, the outdoor air temperature and the staying condition of personnel in the station serve as factors for judging whether shutdown occurs or not; and the phase change cold storage system is adopted to adapt to follow-up load fluctuation caused by passenger flow changes, the downtime of the water chilling unit can be prolonged, and frequent starting and stopping of the water chilling unit are avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of air conditioning systems, and in particular to an energy-saving control method and system for a chiller unit in a subway station. Background Art

[0002] With the rapid development of urban railways, subways are one of the important means of transportation in modern society. According to surveys, the energy consumption of subway air-conditioning systems accounts for 1 / 3-1 / 2 of the total energy consumption during subway operations. Among the energy consumption of air-conditioning systems, the energy consumption of chillers accounts for about 1 / 2. Traditional air-conditioning system control methods are mostly based on experience or fixed operating modes. During non-peak passenger flow periods, the temperature in the station is often too low, resulting in energy waste.

[0003] In order to achieve energy saving in the subway environmental control system, a method of controlling the air conditioning system according to the predicted cooling load has been proposed, and the operating parameters of the chiller are adjusted in real time to achieve the purpose of energy saving. Subway stations are generally equipped with multiple chillers. In the prior art, it can be determined whether one or more chillers can be shut down according to the predicted cooling load. However, the cooling load in the station is easily affected by the passenger flow. Since the passenger flow varies greatly and is difficult to predict, shutting down the chiller may lead to insufficient cooling due to the increase in passenger flow. If the chiller is turned on again, the chiller will be frequently started and stopped, which will increase energy consumption and reduce the service life of the chiller. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for energy-saving control of a chiller in a subway station to improve the above-mentioned problem. In order to achieve the above-mentioned purpose, the technical solution adopted by the present invention is as follows:

[0005] In a first aspect, the present application provides an energy-saving control method for a chiller unit in a subway station, comprising:

[0006] Obtain passenger flow data, indoor and outdoor environmental data and equipment data, perform load forecasting based on the passenger flow data, indoor and outdoor environmental data and equipment data, generate equipment operating parameters according to the load forecasting results, and maintain the indoor temperature within a target range; the passenger flow data includes personnel retention data;

[0007] In a preset time period, the forecast temperature data is obtained, and it is determined whether to shut down at least one chiller according to the forecast temperature data and the personnel retention data. If the chiller is shut down, the phase change cold storage system is used for auxiliary cooling;

[0008] The equipment operation parameters are regenerated according to the equipment data after the auxiliary cooling, the passenger flow data and the indoor and outdoor environmental data.

[0009] In a second aspect, the present application also provides an energy-saving control system for a chiller unit in a subway station, comprising:

[0010] A prediction control module is used to obtain passenger flow data, indoor and outdoor environmental data and equipment data, perform load prediction based on the passenger flow data, indoor and outdoor environmental data and equipment data, generate equipment operating parameters according to the load prediction results, and maintain the indoor temperature within a target range; the passenger flow data includes personnel retention data;

[0011] The start-stop control module is used to obtain the forecast temperature data within a preset time period, and determine whether to shut down at least one chiller according to the forecast temperature data and the personnel retention data. If the chiller is shut down, the phase change cold storage system is used for auxiliary cooling;

[0012] The parameter updating module is used to regenerate the equipment operation parameters according to the equipment data after the auxiliary cooling, the passenger flow data and the indoor and outdoor environmental data.

[0013] The beneficial effects of the present invention are:

[0014] The method of the present application no longer determines whether to shut down the chiller based on the predicted cooling load like the traditional method, but makes a judgment based on the outdoor temperature and the retention of people in the station; the future outdoor temperature and the retention of people can be predicted relatively accurately; and the passenger flow data of those who only stay in the station for a short time is not considered as a factor in shutting down the chiller. When the present application determines that a chiller can be shut down, the phase change cold storage system is turned on for auxiliary cooling, and the load fluctuations caused by subsequent changes in passenger flow are adapted to by controlling the cooling capacity of the phase change cold storage system. The energy consumption of the phase change cold storage system is significantly lower than that of the chiller, which can extend the downtime of the chiller while ensuring the cooling effect in the station, and avoid frequent start and stop of the chiller, thus achieving energy-saving control of the chiller.

[0015] Other features and advantages of the present invention will be set forth in the following description, and in part will become apparent from the description, or may be understood by practicing embodiments of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 It is a schematic flow chart of the energy-saving control method of the chiller unit in a subway station described in an embodiment of the present invention;

[0018] Figure 2It is a schematic structural diagram of the energy-saving control system of the chiller unit in a subway station described in an embodiment of the present invention.

[0019] Explanation of symbols: 100, prediction control module; 110, retention data acquisition unit; 111, first acquisition unit; 112, second acquisition unit; 113, first identification unit; 114, third acquisition unit; 120, station hall load prediction unit; 130, platform load prediction unit; 140, total load calculation unit; 200, start-stop control module; 300, parameter update module. DETAILED DESCRIPTION

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by those skilled in the art to which the present application belongs. The terms used herein in the specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0021] In the present application, the technical features described in an open manner include closed technical solutions composed of the listed features, and also include open technical solutions containing the listed features.

[0022] The factors that have the greatest impact on station temperature are outdoor temperature and passenger flow. Traditional load forecasting methods mostly build forecasting models based on historical data on the environment and passenger flow to predict future loads. However, passenger flow during non-peak commuting hours is actually a factor that varies greatly and is difficult to predict accurately. The current load change may also be quite different from the historical load change. Therefore, after shutting down the chiller according to the predicted load, if the passenger flow increases, insufficient cooling may occur and the chiller needs to be turned on again, resulting in frequent start and stop of the chiller and poor energy-saving effect.

[0023] like Figure 1 As shown, the present application provides an energy-saving control method for a chiller unit in a subway station, comprising step S100, step S200 and step S300.

[0024] Step S100, obtaining passenger flow data, indoor and outdoor environmental data and equipment data, performing load forecasting based on the passenger flow data, indoor and outdoor environmental data and equipment data, generating equipment operating parameters according to the load forecasting results, and maintaining the indoor temperature within a target range; the passenger flow data includes personnel retention data;

[0025] This step is the basic control method of the chiller. During the entire working period, the chiller operates according to the above method.

[0026] The indoor and outdoor environmental data include outdoor dry-bulb temperature, outdoor wet-bulb temperature, outdoor relative humidity, CO2 concentration, temperature and humidity at the concourse and platform levels, etc.;

[0027] The equipment data includes chilled water temperature difference, return air temperature difference, cooling water temperature difference and number of units in operation;

[0028] The passenger flow data includes the number of people entering and leaving the station acquired by the image acquisition device, and also includes the number of people entering and leaving the station recorded by the gate;

[0029] This step uses a pre-trained load prediction model for load prediction. The load prediction model is preferably constructed based on an LSTM network, with historical passenger flow data, equipment parameter data, and environmental data as training inputs and the corresponding historical cooling load as training outputs to obtain a prediction model. Based on the predicted load, the chilled water flow and temperature difference that meet the load can be calculated, and the control strategy can be optimized under a variety of constraints, including dynamically adjusting the control parameters of the chiller according to the set chilled water temperature range, the maximum output and minimum output of each device, etc., to maintain the indoor temperature within the target range.

[0030] In this step, different load prediction models are established for the hall layer and the platform layer respectively. The main difference between the hall layer load prediction model and the platform layer load prediction model is that the passenger flow data are considered differently. In addition, since the platform layer is deeper than the platform layer, its temperature fluctuation is smaller and it is less affected by the outdoor temperature. Therefore, establishing different prediction models for the hall layer and the platform layer can make the prediction value more accurate.

[0031] As an optional implementation manner, passenger flow data, indoor and outdoor environmental data, and equipment data are obtained, and load forecasting is performed based on the passenger flow data, indoor and outdoor environmental data, and equipment data, including:

[0032] Step S110: Use monitoring equipment to collect images of the station hall layer, obtain first passenger flow data according to the image collection results, and use face recognition technology and behavior recognition technology to obtain the data of people staying in the station;

[0033] Specifically, according to the surveillance video of the subway entrance and exit, a passenger flow counting system based on visual recognition is used to obtain the number of people entering and leaving the station, and the number of people in the current station is calculated based on the data as the first passenger flow data;

[0034] The facial recognition technology and behavior recognition technology are used to obtain the data of people staying in the station, specifically:

[0035] Use face recognition technology to process all images collected at the station hall level, obtain facial information of everyone at the station hall level, remove facial information belonging to subway staff, obtain valid facial information, generate a corresponding ID based on each valid facial information, store it in the cache database, and record its corresponding storage time;

[0036] Based on the effective facial information, behavior recognition technology is used to obtain the behavior information of the person, and obtain their location and stay time, as follows:

[0037] When obtaining valid facial information, if the corresponding ID is retrieved and already exists in the cache database, it will not be stored again, and the difference between the storage time of the ID and the current time is calculated to obtain the residence time.

[0038] When the stay time exceeds the preset value, the person is classified as a monitored person, and then the monitored person is tracked and the behavior recognition technology is used to identify the monitored person's behavior information. For example, Openpose's posture recognition technology is used to identify human behaviors such as standing, sitting, walking, and looking at mobile phones. At the same time, face recognition technology can be combined to identify behaviors such as talking, and the behavior information of each monitored person can be obtained.

[0039] Generally, people stay in the station hall for too long because they are resting, cooling off or waiting in the station. Especially in recent years, in order to reduce urban power consumption in summer, more and more city subways have set up cooling off areas for citizens. As a result, there have been people staying in the station hall for a long time. The behaviors of such people are highly overlapping, such as sitting, talking, looking at mobile phones, etc., and their activities are concentrated in a small area. Therefore, the behavioral information of people can be used to predict whether they will stay for a long time.

[0040] While obtaining the behavior information, the location of the monitored person is obtained based on the collected images, and the range of his activities during the current stay time is obtained.

[0041] According to the behavior information, location and current stay time of the monitored personnel, the stranded personnel are identified based on the deep learning network model, the number of stranded personnel is obtained, the facial information of the stranded personnel is stored in the database and the corresponding stay time is recorded, and the stay time is the total time that the person stays in the station hall layer;

[0042] The deep learning network model is pre-trained with a large amount of historical data, specifically including: constructing a GA-BP neural network model, quantifying and assigning values ​​to the behavior information, current stay time, and activity range of the personnel as the input of the model, and taking the total stay time of the personnel as the output, training the GA-BP neural network model to obtain a deep learning network model. The total stay time of the monitored personnel can be predicted through this model, and when the predicted total stay time exceeds the preset value, the person is judged to be a stranded person.

[0043] As a preferred implementation, when the behavior information of a person is quantitatively assigned, the behavior information is firstly classified into features, for example, standing, sitting, walking and running are used as posture features, and each posture feature is assigned a value, such as standing: 0, sitting: 1, walking: 2, running: 3; talking, looking at the phone, answering the phone, etc. are used as behavior features, and each feature is also assigned a value, such as talking: 4, looking at the phone: 5, answering the phone: 6, and the posture features and behavior features that occur at the same time are combined, such as sitting and talking at the same time, that is, the feature combination 14 is obtained, and the feature combination of the person is obtained at a preset time interval (such as 1 minute), such as obtaining five feature combinations to obtain [04, 15, 15, 15, 05], and the time series data is used as the model input. In other embodiments, more feature classifications can be expanded, such as taking the age or expression of the person as a feature and combining it with other features.

[0044] The facial information and ID of each stranded person are stored in the stranded database, and the corresponding actual stranded time is recorded.

[0045] According to the effective facial information of the stranded person, the corresponding historical stranded time is retrieved in the stranded database.

[0046] If the ID of the stranded person is a newly added ID in the stranded database, that is, there is no historical data of his / her stranded time in the stranded database, the total predicted stay time will be used as his / her current stranded time for load forecasting; if there is historical data of his / her stranded time in the stranded database, the mean will be calculated by using the historical data of his / her stranded time and the total predicted stay time, and the calculated result will be used as his / her current stranded time for load forecasting;

[0047] Step S120, predicting the load of the station hall layer according to the environmental data, the first passenger flow data and the personnel retention data;

[0048] This step constructs a station hall load prediction model, which is trained with historical environmental data, first passenger flow data and personnel retention data as input and historical cooling load of the station hall as output. The retention data includes the number of stranded personnel and the retention time of each stranded personnel. The station hall load is obtained through the station hall load prediction model.

[0049] Step S130, predicting the platform load according to the environmental data and the second passenger flow data acquired by the gate system;

[0050] The second passenger flow data obtained by the gate system is the entry data recorded by the gate system. The passenger flow reflected by the data does not include stranded passengers, but passengers traveling normally. Therefore, the second passenger flow data is the actual passenger flow that affects the platform temperature.

[0051] Therefore, the hall load prediction model constructed in this step is trained with historical environmental data and the second passenger flow data as input and historical platform cooling load as output. The platform load is obtained through the platform load prediction model.

[0052] Step S140: Calculate the total load of the subway station according to the hall load and the platform load. That is, the total load of the station can be obtained by calculating the sum of the hall load and the platform load.

[0053] Step S200: obtaining forecast temperature data within a preset time period, and determining whether to shut down at least one chiller according to the forecast temperature data and the personnel retention data; if the chiller is shut down, using a phase change cold storage system for auxiliary cooling; the preset time period is a non-peak passenger flow period;

[0054] Specifically, judging whether to shut down at least one chiller according to the forecast temperature data and the personnel retention data includes:

[0055] In a preset time period, the estimated difference between the indoor temperature and the forecast temperature at at least one subsequent time point is calculated; in this embodiment, the indoor temperature uses the average of the station hall temperature and the platform temperature, and selects the temperature at the next two time points (each time point is 1 hour apart). For example, the current time is 9:20, and the subsequent time points are 10:00 and 11:00, respectively. The estimated difference between the indoor temperature and the forecast temperature at the second subsequent time point, i.e., 11:00, is calculated to be 5.6°C. Preferably, a threshold value can be set in advance. When the estimated difference value is higher than the threshold value, it means that the outdoor temperature will be very high. At this time, the calculation and judgment of the subsequent unit demand value will no longer be performed, and the chiller will not be shut down. Alternatively, the number of chillers that need to be shut down can be judged according to the preset threshold value.

[0056] The detention impact is calculated based on the number of stranded personnel and the corresponding detention time. Specifically, first, the number of historical detention times (i.e., historical detention times) of the person is queried in the detention database based on the person's ID, and a weight is generated for each person based on the person's historical detention times. The more detention times, the greater the weight.

[0057] Multiply each person's weight by their current detention time, and then calculate the total for all people to get the detention impact.

[0058] The unit demand value is calculated based on the estimated difference and the retention impact; the calculation method of the unit demand value is:

[0059] During off-peak hours, we selected two hours with the smallest passenger flow fluctuations, counted the total load value, detention impact, and maximum indoor and outdoor temperature difference of the station hall within two hours, and collected multiple sets of data as samples;

[0060] With the detention influence and the maximum temperature difference between indoor and outdoor as independent variables, and the total load value of the station hall layer as the dependent variable, Matlab was used for nonlinear fitting to obtain the fitting function, which was then used as the calculation formula for the unit demand value. The estimated difference and the detention influence were substituted into the formula to calculate the unit demand value.

[0061] If the unit demand value is less than the preset shutdown threshold, at least one chiller is shut down.

[0062] After the calculation formula of the unit demand value is preliminarily determined, a reasonable judgment on whether to shut down can be achieved by adjusting the shutdown threshold. In general, the shutdown judgment condition of the present application method is more relaxed than the traditional judgment method, which means that the shutdown time can be extended.

[0063] The auxiliary cooling using the phase change cold storage system comprises:

[0064] Based on the forecast temperature data, the auxiliary cooling time for the day is predicted; specifically, the time period suitable for using auxiliary cooling can be estimated based on the forecast temperature, combined with historical daily auxiliary cooling time predictions.

[0065] The maximum flow rate of the phase change material is calculated based on the auxiliary cooling time and the total amount of cold storage in the phase change cold storage system; the total amount of cold storage is the volume of the phase change material that has stored cold in the current phase change system, and the total amount of cold storage may be different every day.

[0066] Maximum flow rate = total amount of cold storage / auxiliary cooling time;

[0067] In order to adjust the auxiliary cooling strategy in real time according to the current load and avoid excessive energy consumption of the chiller that is still in operation as much as possible, firstly, a suitable target temperature is set for the chilled water return according to the selection of the chiller; the difference between the temperature of the chilled water return and the preset target temperature is calculated, and if the difference is less than the preset first threshold, the flow rate of the phase change material is calculated according to the difference; the setting principle of the first threshold is: when the difference is less than the first threshold, the flow rate of the phase change material is adjusted to be sufficient to reduce the chilled water return temperature to the target temperature;

[0068] If the difference is greater than or equal to the preset first threshold, the maximum flow rate is used as the flow rate of the phase change material; at this time, the chilled water return cannot be reduced to the target temperature through the phase change material, and the working energy consumption of the chiller will increase slightly, but the pre-cooling effect of the phase change material also greatly alleviates the working pressure of the chiller.

[0069] The formula for calculating the flow rate of the phase change material based on the difference is:

[0070]

[0071] Among them, L 相 is the flow rate of phase change material, L 水 is the chilled water flow rate, ρ 水 is the density of frozen water, c 水 is the constant pressure specific heat of chilled water, Δt is the difference between the return temperature of chilled water and the preset target temperature, λ is the phase change heat transfer coefficient, K is the unit cold storage capacity of the phase change material (kJ / m 3 );

[0072] The K value is determined according to the selected phase change material, and the λ value is related to factors such as the selection of the heat exchanger and the type of phase change material, and can be determined through multiple experiments.

[0073] To implement this method, a heat exchanger needs to be set on the chilled water return pipeline. The phase change material cools the chilled water return in the heat exchanger. The cooled chilled water return enters the evaporator, and the phase change material after absorbing heat is sent to the material storage tank for storage.

[0074] As an optional implementation, the material storage tank includes a plurality of storage tank units arranged in parallel, each storage tank unit is provided with a temperature sensor, and the cold storage method of the phase change cold storage system includes:

[0075] Obtain the chilled water flow. When the system is in low-load operation, the unit will use a frequency converter to adjust the chilled water flow to reduce energy consumption, but the chilled water flow cannot be lower than the minimum value allowed by the unit. Therefore, when the obtained chilled water flow approaches the minimum allowed value, for example, when it is lower than the preset second threshold, the operating frequency of the chilled water pump is kept constant or unchanged, and part of the chilled water outlet is sent to the material storage tank, and the phase change material of each storage tank unit is cooled down in turn for cold storage. When the operating frequency of the water pump rises to a certain value or the chilled water return temperature rises to a certain value, the cold storage ends and the chilled water returns to normal water discharge.

[0076] During periods when passenger flow is far from the design value, the temperature in the subway station is generally low. This method uses the cooling capacity during low-load periods to store cold for the phase change material, making reasonable use of energy. The daily low-load operation time may be different, so the daily phase change material cold storage capacity is also different. To further ensure that the cold storage capacity is sufficient, cold storage can be carried out at night when the electricity price is low.

[0077] The phase change material is obtained by dispersing phase change capsule material or phase change metal particles in a flow carrier, so the phase change material can be transported through a pipeline. When auxiliary cooling is provided, the phase change material is output from the cold storage tank unit and heat exchange is performed with the chilled water in a convection form.

[0078] Step S300: regenerate equipment operating parameters according to the equipment data after auxiliary cooling, passenger flow data, and indoor and outdoor environmental data.

[0079] Measure the chilled water return temperature after phase change material cooling, use this temperature to update the chilled water temperature difference, and update the number of running units at the same time. Use passenger flow data, indoor and outdoor environmental data and updated equipment data to perform load forecasting and regenerate equipment operating parameters.

[0080] On the other hand, the present application also provides an energy-saving control system for a chiller in a subway station, see Figure 2 ,include:

[0081] The prediction control module 100 is used to obtain passenger flow data, indoor and outdoor environmental data and equipment data, perform load prediction based on the passenger flow data, indoor and outdoor environmental data and equipment data, generate equipment operating parameters according to the load prediction results, and maintain the indoor temperature within the target range; the passenger flow data includes personnel retention data;

[0082] The start-stop control module 200 is used to obtain the forecast temperature data within a preset time period, and determine whether to shut down at least one chiller according to the forecast temperature data and the personnel retention data. If the chiller is shut down, the phase change cold storage system is used for auxiliary cooling;

[0083] The parameter updating module 300 is used to regenerate the equipment operation parameters according to the equipment data after the auxiliary cooling, the passenger flow data and the indoor and outdoor environmental data.

[0084] As an optional implementation, the prediction control module 100 includes:

[0085] The retention data acquisition unit 110 is used to collect images of the station hall layer, obtain the first passenger flow data according to the image collection results, and obtain the retention data of the personnel in the station by using the face recognition technology and the behavior recognition technology;

[0086] The hall load prediction unit 120 is used to predict the hall load according to the environment data, the equipment data, the first passenger flow data and the personnel retention data;

[0087] The platform load prediction unit 130 is used to predict the platform load according to the environment data, the equipment data and the second passenger flow data obtained by the gate system;

[0088] The total load calculation unit 140 is used to calculate the total load of the subway station according to the hall load and the platform load.

[0089] As an optional implementation, the retention data acquisition unit 110 includes:

[0090] The first acquisition unit 111 is used to acquire facial information of people in the station, remove facial information belonging to subway staff, and obtain valid facial information;

[0091] A second acquisition unit 112 is used to obtain the behavior information of the person based on the effective facial information by using the behavior recognition technology, and obtain the person's location and stay time;

[0092] The first recognition unit 113 is used to identify the stranded persons based on the deep learning network model according to the behavior information, location and residence time of the persons, obtain the number of stranded persons, store the facial information of the stranded persons in the database and record the corresponding residence time;

[0093] The third acquisition unit 114 is used to retrieve the corresponding historical detention time of the stranded person in the database according to the effective facial information of the stranded person.

[0094] It should be noted that, regarding the system in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0095] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for energy-saving control of a chiller unit in a subway station, characterized in that: include: Obtain passenger flow data, indoor and outdoor environmental data and equipment data, perform load forecasting based on the passenger flow data, indoor and outdoor environmental data and equipment data, generate equipment operating parameters according to the load forecasting results, and maintain the indoor temperature within a target range; the passenger flow data includes personnel retention data; Obtain forecast temperature data within a preset time period, and determine whether to shut down at least one chiller based on the forecast temperature data and personnel retention data. If the chiller is shut down, use a phase change cold storage system for auxiliary cooling; The equipment operation parameters are regenerated according to the equipment data after the auxiliary cooling, the passenger flow data and the indoor and outdoor environmental data.

2. The energy-saving control method for a chiller unit in a subway station according to claim 1, characterized in that: Obtaining passenger flow data, indoor and outdoor environmental data, and equipment data, and performing load forecasting based on the passenger flow data, indoor and outdoor environmental data, and equipment data, including: Capture images of the station hall, obtain the first passenger flow data based on the image acquisition results, and use face recognition technology and behavior recognition technology to obtain the data on the number of people staying in the station; Predict the load of the concourse level based on environmental data, equipment data, first passenger flow data and personnel retention data; Predict the platform load based on the environmental data, equipment data and the second passenger flow data obtained by the gate system; The total load of the subway station is calculated based on the hall load and the platform load.

3. The energy-saving control method for a chiller unit in a subway station according to claim 2, characterized in that: The facial recognition technology and behavior recognition technology are used to obtain the data of people staying in the station, including: Obtain facial information of people in the station, remove facial information belonging to subway staff, and obtain valid facial information; Based on the effective facial information, behavior recognition technology is used to obtain the behavior information of the person, and the person's location and stay time are obtained; According to the behavior information, location and length of stay of the personnel, the stranded personnel are identified based on the deep learning network model, the number of stranded personnel is obtained, the facial information of the stranded personnel is stored in the database and the corresponding length of stay is recorded; According to the effective facial information of the stranded person, the corresponding historical stranded time is retrieved from the database.

4. The energy-saving control method for a chiller unit in a subway station according to claim 3, characterized in that: The step of judging whether to shut down at least one chiller according to the predicted temperature data and the personnel retention data includes: Calculate the estimated difference between the indoor temperature and the predicted temperature at at least one subsequent time point within a preset time period; According to the number of stranded personnel and the corresponding historical stranded time, the stranded impact is calculated; The unit demand value is calculated based on the estimated difference and the retention impact; If the unit demand value is less than the preset shutdown threshold, at least one chiller is shut down.

5. The energy-saving control method for chillers in subway stations according to claim 1, characterized in that: The auxiliary cooling using the phase change cold storage system comprises: According to the forecast temperature data, the auxiliary cooling time of the day is predicted; Calculating the maximum flow rate of the phase change material according to the auxiliary cooling time and the total amount of cold storage of the phase change cold storage system; Calculate the difference between the temperature of the chilled water return water and the preset target temperature. If the difference is less than the preset second threshold, calculate the flow rate of the phase change material according to the difference; if the difference is greater than or equal to the preset second threshold, use the maximum flow rate as the flow rate of the phase change material; The phase change material is used to cool the return chilled water, and the cooled return chilled water enters the evaporator.

6. The energy-saving control method for chillers in subway stations according to claim 4, characterized in that: According to the number of stranded personnel and the corresponding historical detention time, the detention impact is calculated, including: According to the effective facial information of the detained person, the number of his / her historical detention time is queried in the database to obtain his / her historical detention times; Generating a weight for each stranded person according to the historical number of stranded times; Based on the weight of each stranded person and the historical detention time, the detention impact is obtained.

7. The energy-saving control method for chillers in subway stations according to claim 1, characterized in that: The phase-change cold storage system includes a material storage tank, and the cold storage method of the phase-change cold storage system includes: The chilled water flow is obtained. When the chilled water flow is lower than a preset second threshold, the operating frequency of the chilled water pump is kept unchanged, and part of the chilled water outlet is transported to the material storage tank to cool the phase change material in the material storage tank for cold storage.

8. An energy-saving control system for a chiller unit in a subway station, characterized in that: include: A prediction control module is used to obtain passenger flow data, indoor and outdoor environmental data and equipment data, perform load prediction based on the passenger flow data, indoor and outdoor environmental data and equipment data, generate equipment operating parameters according to the load prediction results, and maintain the indoor temperature within a target range; the passenger flow data includes personnel retention data; The start-stop control module is used to obtain the forecast temperature data within a preset time period, and determine whether to shut down at least one chiller according to the forecast temperature data and the personnel retention data. If the chiller is shut down, the phase change cold storage system is used for auxiliary cooling; The parameter updating module is used to regenerate the equipment operation parameters according to the equipment data after the auxiliary cooling, the passenger flow data and the indoor and outdoor environmental data.

9. The energy-saving control system for chillers in subway stations according to claim 8, characterized in that: The prediction control module comprises: The retention data acquisition unit is used to collect images of the station hall layer, obtain the first passenger flow data according to the image collection results, and obtain the retention data of the personnel in the station by using the face recognition technology and the behavior recognition technology; The station hall load prediction unit is used to predict the station hall load based on the environmental data, equipment data, the first passenger flow data and the personnel retention data; A platform load prediction unit, used for predicting the platform load according to the environment data, the equipment data and the second passenger flow data acquired by the gate system; The total load calculation unit is used to calculate the total load of the subway station according to the hall load and the platform load.

10. The energy-saving control system for chillers in subway stations according to claim 9, characterized in that: The retention data acquisition unit comprises: The first acquisition unit is used to acquire facial information of people in the station, remove facial information belonging to subway staff, and obtain valid facial information; A second acquisition unit is used to obtain the behavior information of the person based on the effective facial information by using behavior recognition technology, and obtain the person's location and stay time; The first recognition unit is used to identify the stranded persons based on the deep learning network model according to the behavior information, location and residence time of the persons, obtain the number of stranded persons, store the facial information of the stranded persons in the database and record the corresponding residence time; The third acquisition unit is used to retrieve the corresponding historical detention time in the database according to the effective facial information of the stranded person.

Images