Online monitoring and early warning method and system for state of air conditioner of metro vehicle

Abstract

The invention discloses a subway vehicle air conditioner state on-line monitoring and early warning method which is characterized by comprising the steps that a host obtains air conditioner state data and sends the air conditioner state data to first monitoring equipment, and the air conditioner state data is obtained by monitoring an air conditioner control panel through state on-line monitoring nodes; the air conditioner state data comprises a unit working state, a unit power supply state, a unit communication state, an air conditioner return air temperature, an air conditioner fresh air temperature, an air conditioner set temperature and air conditioner service time; the intelligent operation and maintenance platform obtains the air conditioner state data as an input data set and generates the residual predicted life of the air conditioner control panel by using a pre-trained residual life prediction model of the air conditioner control panel; and the intelligent operation and maintenance platform generates early warning information according to the residual predicted life of the air conditioner control panel and sends the early warning information to second monitoring equipment. According to the method, macroscopic reasoning and early warning can be realized, and the response time and the accuracy of maintenance are improved and shortened.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of rail transit, in particular to the field of rail transit equipment operation state monitoring. BACKGROUND

[0002] The air conditioning system of a subway vehicle is a key device for ensuring the comfort of the environment in the vehicle cabin, and the performance thereof will directly affect the comfort of passengers and the air quality in the passenger compartment and even the health of passengers. In order to ensure that the air conditioning controller of the subway vehicle operates in good condition, routine maintenance and inspection of the air conditioning system of the train are required. In the prior art, the air conditioning system of the subway uses a PLC as a control core to control the temperature in the vehicle cabin by controlling the compressor, the condenser fan and the ventilator. At present, the maintenance method for the air conditioning of the subway vehicle mainly relies on manual detection, and the air conditioning controller in each vehicle cabin of the subway vehicle needs to be tested to determine whether it operates in good condition. This method has two major problems: first, the number of subway vehicles is large, and the inspection process is complex and the production preparation process is numerous, resulting in low efficiency. The inspection of one air conditioning control system needs to complete six processes of opening the monitoring software, opening the air conditioning control cabinet, connecting the communication line, downloading the data, storing the computer and checking the fault data, and the efficiency is very low by relying on manual testing. Second, the maintenance personnel need to connect the PLC module of the air conditioning controller with the test computer through a data line, and then test it by using the maintenance test software. The communication interface is mechanically damaged due to frequent plugging, which may cause abnormal data and affect the test results. Moreover, the version of the maintenance test software is old and cannot be compatible with the current computer operating system, which greatly affects the maintenance work. SUMMARY

[0003] To solve the above problems, the purpose of the present application is to provide a subway vehicle air conditioning state online monitoring and early warning method and device which can realize macro reasoning and early warning and improve the detection efficiency.

[0004] The present application provides a subway vehicle air conditioning state online monitoring and early warning method, which comprises the following steps:

[0005] S1) The host obtains air conditioning state data and sends it to the first monitoring device, wherein the air conditioning state data is obtained by monitoring the air conditioning control board through the state online monitoring node, and the air conditioning state data comprises the working state of the unit, the power supply state of the unit, the communication state of the unit, the air conditioning return air temperature, the air conditioning fresh air temperature, the air conditioning set temperature and the air conditioning service time;

[0006] S2) The intelligent operation and maintenance platform obtains the air conditioning state data as an input data set and generates the remaining predicted life of the air conditioning control board by using a pre-trained air conditioning control board remaining life prediction model;

[0007] S3) The intelligent operation and maintenance platform generates early warning information according to the remaining predicted life of the air conditioner control panel and sends the early warning information to the second monitoring device.

[0008] Further, the subway vehicle air conditioner state online monitoring and early warning method, the unit working state includes control panel overall function area state, central control panel function area state, power supply function area state, communication function area state, temperature acquisition function area state, output control function area state and state detection function area state.

[0009] Further, the subway vehicle air conditioner state online monitoring and early warning method, the control model of the control panel overall function area state includes:

[0010] F 控制板 =F 中央控制 +F 供电 +F 通讯 +F 温度采集 +F 输出控制 +F 状态监测 ;

[0011] F 中央控制 =f 中央控制 (x1,x2,...,x n );

[0012] F 供电 =f 供电 (y1,y2,...,y n );

[0013] F 通讯 =f 通讯 (z1,z2,...,z n );

[0014] F 温度采集 =f 温度采集 (m1,m2,...,m n );

[0015] F 输出控制 =f 输出控制 (n1,n2,...,n n );

[0016] F 状态检测 =f 状态检测 (p1,p2,...,p n );

[0017] Wherein, F 控制板 is the control panel overall function area state;

[0018] F 中央控制 is the central control panel function area state, x1, x2,..., x nThe state of the central control board function area includes the states of the components and devices;

[0019] F 供电 The state of the power supply function area is y1, y2,..., y n The state of the power supply function area includes the states of the components and devices;

[0020] F 通讯 The state of the communication function area is z1, z2,..., z n The state of the communication function area includes the states of the components and devices;

[0021] F 温度采集 The state of the temperature acquisition function area is m1, m2,..., m n The state of the temperature acquisition function area includes the states of the components and devices;

[0022] F 输出控制 The state of the output control function area is n1, n2,..., n n The state of the output control function area includes the states of the components and devices;

[0023] F 状态监测 The state of the state detection function area is p1, p2,..., p n The state of the state detection function area includes the states of the components and devices.

[0024] Further, the subway vehicle air conditioner state online monitoring and early warning method, the remaining life prediction model of the air conditioner control board adopts a CNN-LSTM prediction model, and the pre-training method of the air conditioner control board remaining life prediction model includes:

[0025] Obtain air conditioner control board life data, which is obtained according to the performance parameters of the key components and devices of the air conditioner control board and related maintenance records;

[0026] The air conditioner state data and the air conditioner control board life data are normalized and divided into a training data set and a test data set;

[0027] The training data set and the test data set are used to train the air conditioner control board remaining life prediction model and are corrected using performance evaluation, thereby obtaining the pre-trained air conditioner control board remaining life prediction model.

[0028] Further, the subway vehicle air conditioner state online monitoring and early warning method, the normalized calculation expression is:

[0029]

[0030] Wherein, X * represents the normalized data;

[0031] X represents the input data set;

[0032] X max represents the maximum value in the input data set;

[0033] X min represents the minimum value in the input data set.

[0034] Further, the subway vehicle air conditioning state online monitoring and early warning method, the performance evaluation uses root mean square error RMSE and decision coefficient R 2 for evaluation, the calculation expression includes:

[0035]

[0036]

[0037] wherein N represents the total number of samples of the test data set;

[0038] represents the predicted value of the remaining life of the data sample of the ith test data set;

[0039] represents the true value of the data sample of the ith test data set;

[0040] RUL mean represents the average value of the actual value of the remaining useful life of the total number of samples of the test data set.

[0041] Further, the subway vehicle air conditioning state online monitoring and early warning method, the step S3) further includes:

[0042] When the remaining predicted life of the air conditioning control panel is greater than or equal to 16 years and less than 20 years, it is judged as low risk, and the early warning information is not sent;

[0043] When the remaining predicted life of the air conditioning control panel is greater than or equal to 9 years and less than 16 years, it is judged as medium risk, and the early warning information is sent to the first monitoring device;

[0044] When the remaining predicted life of the air conditioning control panel is greater than or equal to 0 and less than 9 years, it is judged as high risk, and the early warning information is sent to the monitoring device.

[0045] Further, the subway vehicle air conditioner state online monitoring and early warning method, the parameter setting of the CNN-LSTM prediction model includes CNN network model parameter setting and LSTM network model parameter setting, the CNN network model parameter setting includes CNN activation function, network convolution layer and pooling operation, the LSTM network model parameter setting includes LSTM network layer, LSTM activation function, loss function and optimizer setting.

[0046] Further, the subway vehicle air conditioner state online monitoring and early warning method, the CNN activation function uses Relu activation function, the number of network convolution layers is two layers, the pooling operation is the maximum pooling function, the LSTM network layer is two layers, the LSTM activation function is Tanh activation function, the loss function is mean square error loss function, and the optimizer is adma optimizer.

[0047] The application also discloses a subway vehicle air conditioner state online monitoring and early warning system using the subway vehicle air conditioner state online monitoring and early warning method.

[0048] The state online monitoring node is used to acquire air conditioner state data, and the air conditioner state data includes unit working state, unit power supply state, unit communication state, air conditioner return air temperature, air conditioner fresh air temperature, air conditioner set temperature and air conditioner service time.

[0049] The intelligent operation and maintenance platform is used to take the air conditioner state data as input and generate air conditioner control panel remaining predicted life by using a pre-trained air conditioner control panel remaining life prediction model, and is used to generate early warning information according to the air conditioner control panel remaining predicted life and send the early warning information to the second monitoring device.

[0050] The host is used to receive the air conditioner state data acquired by the state online monitoring node and transmit the air conditioner state data to the intelligent operation and maintenance platform through a bus, and is used to send the air conditioner state data to the first monitoring device.

[0051] The technical scheme provided by the application embodiment has the following advantages:

[0052] 1. Since the state online monitoring node is adopted, the system can be embedded at the air conditioner controller communication interface, the existing system sensing data is used, the volume and cost of the monitoring system are saved, and engineering modification is facilitated.

[0053] 2. Since the acquired air conditioner state data is sent to the first monitoring device, the accuracy of maintenance work is improved.

[0054] 3. Macro-inference is realized by using the pre-trained air conditioner control panel residual life prediction model to complete the residual service life prediction.

[0055] 4. The response time of maintenance is shortened by sending early warning information to the second monitoring device. BRIEF DESCRIPTION OF DRAWINGS

[0056] The accompanying drawings, which form a part of this application, are included to provide a further understanding of the application and are incorporated in and constitute a part of this application. The drawings illustrate embodiments of the present application and, together with the description, serve to explain the principles of the present application.

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and all other drawings obtained by those of ordinary skill in the art without creative labor based on these drawings are within the protection scope of the present application.

[0058] In addition, the drawings are not drawn in a 1:1 ratio, and the relative sizes of the various elements are only exemplarily drawn in the drawings, but not necessarily drawn in true proportions. In the drawings:

[0059] Figure 1 The preferred subway vehicle air conditioner state online monitoring and early warning method flowchart of the embodiment of the present application;

[0060] Figure 2 The preferred CNN-LSTM prediction model flowchart of the embodiment of the present application;

[0061] Figure 3 The preferred subway vehicle air conditioner state online monitoring and early warning system of the embodiment of the present application;

[0062] Figure 4 The residual service life prediction effect diagram of the specific application embodiment of the preferred subway vehicle air conditioner state online monitoring and early warning method and system of the embodiment of the present application. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solutions and advantages of the present application clearer, the technical solutions in the embodiments of the present application will be described in more detail below in combination with the drawings of the embodiments of the present application. In the drawings, the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions. The described embodiments are part of the embodiments of the present application, but not all the embodiments. The embodiments described below by referring to the drawings are exemplary and are intended to explain the present application, but cannot be understood as limiting the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor are within the protection scope of the present application.

[0064] Further, it needs to be explained that, unless otherwise explicitly specified and limited, the "installation", "connection", "linkage" and similar words used in the description of the present application should be understood in a broad sense, for example, the connection can be fixed connection, or detachable connection, or integral connection; can be mechanical connection, or electrical connection; can be direct connection, or indirect connection through intermediate medium, or internal connection of two elements, and the person skilled in the art can understand the specific meaning of the words in the present application according to the specific circumstances.

[0065] Figure 1 The preferred subway vehicle air conditioner state online monitoring and early warning method flow chart of the embodiment of the present application is shown in Figure Figure 1 The subway vehicle air conditioner state online monitoring and early warning method, characterized in that, comprises:

[0066] S1) The host obtains air conditioner state data and sends it to the first monitoring device, the air conditioner state data is obtained by the state online monitoring node monitoring the air conditioner control panel, and the air conditioner state data includes unit working state, unit power supply state, unit communication state, air conditioner return air temperature, air conditioner fresh air temperature, air conditioner set temperature, air conditioner service time;

[0067] S2) The intelligent operation and maintenance platform obtains the air conditioner state data as input data set and uses the pre-trained air conditioner control panel remaining life prediction model to generate air conditioner control panel remaining prediction life;

[0068] S3) The intelligent operation and maintenance platform generates early warning information according to the air conditioner control panel remaining prediction life and sends it to the second monitoring device.

[0069] Preferably, the state online monitoring node collects air conditioner state data as input data set for the input data set X of the remaining service life prediction model through the communication bus, and the relationship is as follows:

[0070] X=(X1,X2,X3...X n );

[0071] X n =W n +E n +C n +(R T ) n +(P T ) n +(S T ) n +T n ;

[0072] Wherein, n is the state online monitoring node ID number, W is the unit working state, E is the unit power supply state, C is the unit communication state, RT is the air conditioner return air temperature, PT is the air conditioner fresh air temperature, ST is the air conditioner set temperature, T is the service time.

[0073] Preferably, the unit working state includes the control board overall function area state, the central control board function area state, the power supply function area state, the communication function area state, the temperature acquisition function area state, the output control function area state and the state detection function area state.

[0074] The state of each functional layer of the air conditioner controller is reflected by the state of the components contained in each functional area: when the components in the functional area fail, the corresponding functional area fails, and the control model is as follows:

[0075] F 控制板 = F 中央控制 + F 供电 + F 通讯 + F 温度采集 + F 输出控制 + F 状态监测 ;

[0076] F 中央控制 = f 中央控制 (x1,x2,...,x n );

[0077] F 供电 = f 供电 (y1,y2,...,y n );

[0078] F 通讯 = f 通讯 (z1,z2,...,z n );

[0079] F 温度采集 = f 温度采集 (m1,m2,...,m n );

[0080] F 输出控制 = f 输出控制 (n1,n2,...,n n );

[0081] F 状态检测 = f 状态检测 (p1,p2,...,p n );

[0082] Wherein, F 控制板 is the control board overall function area state;

[0083] F 中央控制x1, x2,..., x for the central control panel function area state n x1, x2,..., x for the component state contained in the central control panel function area;

[0084] F 供电 y1, y2,..., y for the power supply function area state n y1, y2,..., y for the component state contained in the power supply function area;

[0085] F 通讯 z1, z2,..., z for the communication function area state n z1, z2,..., z for the component state contained in the communication function area;

[0086] F 温度采集 m1, m2,..., m for the temperature acquisition function area state n m1, m2,..., m for the component state contained in the temperature acquisition function area;

[0087] F 输出控制 n1, n2,..., n for the output control function area state n n1, n2,..., n for the component state contained in the output control function area;

[0088] F 状态监测 p1, p2,..., p for the state detection function area state n p1, p2,..., p for the component state contained in the state detection function area.

[0089] Preferably, the air conditioner control panel remaining life prediction model adopts a CNN-LSTM prediction model, which combines a Convolutional Neural Network (CNN) and a Long Short-Term Memory (LSTM) network for processing sequence data and making predictions, and the pre-training method of the air conditioner control panel remaining life prediction model comprises:

[0090] Obtaining air conditioner control panel life data, which is obtained according to the performance parameters of key components of the air conditioner control panel and related maintenance records;

[0091] Dividing the normalized air conditioner state data and air conditioner control panel life data into a training data set and a test data set as input;

[0092] Training the air conditioner control panel remaining life prediction model using the training data set and the test data set and correcting it using performance evaluation to obtain the pre-trained air conditioner control panel remaining life prediction model.

[0093] Specifically, the complete data set of the prediction model can be constructed by collecting the performance parameters of key components on the air conditioning control panel and the related maintenance records and other data of the air conditioning control panel, and then the CNN-LSTM prediction model is constructed to realize the prediction of the remaining service life of the air conditioning control panel of the subway.

[0094] Preferably, the air conditioning control panel life data can be reflected by the performance of the key components, and the mathematical model is as follows:

[0095]

[0096] Wherein, rul represents the remaining life of the corresponding component, and α, δ, λ, γ, ε represent the weight coefficients corresponding to the component.

[0097] Preferably, the parameter setting of the CNN-LSTM prediction model includes the CNN network model parameter setting and the LSTM network model parameter setting, the CNN network model parameter setting includes the CNN activation function, the network convolution layer and the pooling operation, and the LSTM network model parameter setting includes the LSTM network layer, the LSTM activation function, the loss function and the optimizer setting. Wherein, the CNN activation function uses Relu activation function, ReLU (Rectified Linear Unit) is a commonly used activation function, the number of network convolution layers is two, the pooling operation is the maximum pooling function, the LSTM network layer is two, the LSTM activation function is Tanh (hyperbolic tangent) activation function, the loss function is mean square error loss function, and the optimizer is adma optimizer.

[0098] Specifically, the parameter setting of the CNN-LSTM prediction model: the CNN network model uses the activation function of Relu, two network convolution layers, and the maximum pooling function; the LSTM network model uses two LSTM network layers, the activation function of Tanh, uses the mean square error as the loss function, and adopts the "adam" optimizer for gradient optimization. Adam (Adaptive Moment Estimation) is a commonly used adaptive learning rate optimization algorithm, which combines the ideas of momentum gradient descent and adaptive learning rate adjustment. For details, see Table 1 and Table 2 below:

[0099] Table 1

[0100]

[0101] Table 2

[0102]

[0103] Figure 2 The preferred CNN-LSTM prediction model flow chart of the embodiment of the application is as follows:Figure 2 As shown in the figure, the prediction process of the CNN-LSTM prediction model is as follows: data preprocessing, CNN-LSTM prediction model training, CNN-LSTM prediction model testing, and CNN-LSTM prediction model performance evaluation results.

[0104] The data preprocessing in the prediction process is to take the state data and life data obtained by the state online monitoring node as the data input of the prediction model, normalize the sample data, and then divide it into a training data set and a test data set. The normalization formula is as follows:

[0105]

[0106] wherein, X * represents the normalized data;

[0107] X represents the input data set;

[0108] X max represents the maximum value in the input data set;

[0109] X min represents the minimum value in the input data set.

[0110] The CNN-LSTM prediction model training in the prediction process is to initialize the parameters of the built CNN-LSTM remaining useful life prediction model, then train the remaining useful life prediction model using the divided training data set, stop training when the training index meets the predetermined requirement, and output and save the CNN-LSTM-based remaining useful life prediction model obtained by training.

[0111] The CNN-LSTM prediction model testing in the prediction process is to input the data samples of the test data set into the trained remaining useful life prediction model, output the corresponding remaining useful life prediction value, calculate the error between the remaining useful life prediction value and the true value of the remaining useful life using the performance evaluation index, and realize the performance evaluation of the remaining useful life prediction model. The evaluation index of the prediction model adopts the root mean square error RMSE and the determination coefficient R 2 commonly used in regression problems, and the calculation formula is as follows:

[0112]

[0113]

[0114] wherein, N represents the total number of samples of the test data set;

[0115] represents the prediction value of the remaining life of the i-th data sample of the test data set;

[0116] a true value of a data sample of the ith test data set;

[0117] RUL mean an average value of actual values of the remaining useful life of the total number of samples of the test data set.

[0118] Preferably, the step S3) further comprises:

[0119] When the remaining predicted life of the air conditioner control panel is greater than or equal to 16 years and less than 20 years, it is judged as low risk, and the early warning information is not sent.

[0120] When the remaining predicted life of the air conditioner control panel is greater than or equal to 9 years and less than 16 years, it is judged as medium risk, and the early warning information is sent to the first monitoring device.

[0121] When the remaining predicted life of the air conditioner control panel is greater than or equal to 0 and less than 9 years, it is judged as high risk, and the early warning information is sent to the second monitoring device. It is worth noting that the preferred first monitoring device and second monitoring device of the embodiment include handheld monitoring devices, mobile phones, computers and other devices that can receive device status.

[0122] Specifically, the early warning information contains an early warning parameter rul 空调 , an early warning formula is established, and the formula is as follows:

[0123] rul 空调 = min[rul(x1),rul(y1,y2,y3,y8),rul(m7),rul(p1,p2,p3),rul(n1,n2,n3),rul(z 14 )];

[0124] According to the RUL prediction value, a performance evaluation index is established, and the performance evaluation result of the CNN-LSTM prediction model can be previewed. Preferably, the design service life of the subway air conditioner controller is 20 years, and three risk evaluation levels are set: when 16≤rul 空调 <20, it is low risk, and no early warning information is sent; when 9≤rul 空调 <16, it is medium risk, and a prompt information is sent; and when 0≤rul 空调 <9, it is high risk, and an early warning information is sent.

[0125] Figure 3 The subway vehicle air conditioner state online monitoring and early warning system is preferred for the embodiment of the present application. As Figure 2 shown, the subway vehicle air conditioner state online monitoring and early warning system disclosed by the present application uses the subway vehicle air conditioner state online monitoring and early warning method as described above, and comprises:

[0126] a state online monitoring node 1 configured to acquire air conditioner state data, the air conditioner state data including a unit working state, a unit power supply state, a unit communication state, an air conditioner return air temperature, an air conditioner fresh air temperature, an air conditioner set temperature, and an air conditioner service time;

[0127] an intelligent operation and maintenance platform 2 configured to use the air conditioner state data as input and generate an air conditioner control panel remaining predicted life using a pre-trained air conditioner control panel remaining life prediction model, and configured to generate early warning information according to the air conditioner control panel remaining predicted life and send the early warning information to the second monitoring device 3;

[0128] a host 4 configured to receive the air conditioner state data acquired by the state online monitoring node and transmit the air conditioner state data to the intelligent operation and maintenance platform through a bus, and configured to send the air conditioner state data to the first monitoring device 5.

[0129] Specifically, the state online monitoring node 1 is connected with an air conditioner controller B of each car of the metro vehicle through a communication bus A to collect real-time air conditioner state data. The state online monitoring node is embedded at a communication interface of the air conditioner controller to utilize the sensing data of the existing system, thereby saving the volume and cost of the monitoring system and facilitating engineering modification.

[0130] The host 4 is divided into two ends, one end receives the data collected by the state online monitoring node 1 of each car through wireless radio frequency technology, and the other end is connected to a train-ground transmission system 6 through an MVB bus C and synchronizes the state data to the first monitoring device 5 (such as a handheld monitoring device) for use by maintenance personnel during work. In the present embodiment, the data is sent to the handheld detection device through wireless radio frequency technology, thereby improving the accuracy of maintenance work.

[0131] The intelligent operation and maintenance platform 2 is connected to the train-ground transmission system 6 through a private network to perform macroscopic identification on the health state of the air conditioner system, use a CNN-LSTM prediction model to perform life evaluation, form early warning information, and send the early warning information to the second monitoring device (for example, send the early warning information to maintenance personnel in the form of a short message). The present application uses the CNN-LSTM prediction model to realize macro reasoning and complete remaining useful life prediction through a complete data set collected in the early stage.

[0132] Figure 4 The present application is a preferred embodiment of the metro vehicle air conditioner state online monitoring and early warning method and system. The remaining useful life prediction effect diagram is as follows: Figure 3 and Figure 4As shown, the subway vehicle air conditioner state online monitoring and early warning system realizes real-time collection of air conditioner control state data during subway vehicle operation, realizes data interconnection of the whole train level by using wireless network, and the system functions include uninterrupted collection of subway air conditioner controller state data, real-time data transmission, communication wireless transmission, intelligent operation and maintenance platform and handheld detection equipment synchronous information, abnormal data alarm and other functions.Curve I represents the true value, and curve II represents the predicted value, from Figure 4 It can be seen from the above that the preferred subway vehicle air conditioner state online monitoring and early warning method and system of the embodiment obtains a model with good prediction effect by training the remaining service life prediction model of CNN-LSTM. The remaining service life prediction value of the model is relatively close to the true value of the remaining service life, and the prediction effect is good at each stage.

[0133] The present application utilizes digitalization and artificial intelligence technology to provide a subway vehicle air conditioner state online monitoring and early warning method and system, which mainly realizes the following functions: 1) realizing online digital perception of the state of the subway air conditioning system, collecting air conditioner control state data in real time during subway vehicle operation, and realizing data interconnection of the whole train level by using wireless network; 2) realizing macro recognition of the health state of the air conditioning system, establishing a digital twin model through hierarchical coupling factor analysis of air conditioning system state perception data, forming a complete feature data set, using CNN-LSTM network power to build a macro recognition of air conditioning state in the cloud, and sending the warning result to the maintenance personnel through the intelligent operation and maintenance platform.

[0134] The present application has the following advantages: 1) embedded miniaturization, by installing an embedded monitoring device at the communication interface of the air conditioner controller, the existing system sensor data can be effectively utilized, the volume and cost of the monitoring system are saved, and the engineering modification characteristics are good; 2) online digital perception, through automatic monitoring means, the original complex detection process is realized, the state data is online digitized, and the enterprise digital empowerment is improved; 3) having global state recognition ability, by constructing a digital twin model and establishing a complete feature data set, macro reasoning and early warning are realized, and the detection efficiency is improved.

[0135] Compared with the prior art, the present application has the following advantages:

[0136] Firstly, the present application installs an embedded monitoring device at the communication interface of the air conditioner controller, utilizes the existing system sensor data, saves the volume and cost of the monitoring system, and has good engineering modification characteristics.

[0137] Secondly, the present application sends the state information of the air conditioner controller collected by the wireless radio frequency technology to the handheld detection equipment, improves the accuracy of maintenance work.

[0138] Thirdly, the application utilizes a CNN-LSTM prediction model to realize macro reasoning and complete remaining useful life prediction through collecting complete data sets in advance.

[0139] Fourthly, the application establishes an early warning formula, evaluates early warning information, generates early warning short messages, and improves the work efficiency of maintenance personnel.

[0140] Those skilled in the art will appreciate that information, signals, and data can be represented using any of a variety of different technologies and techniques. For example, data, instructions, commands, information, signals, bits, symbols, and chips that can be referenced throughout the above description can be represented by voltages, currents, electromagnetic waves, magnetic fields or particles, optical fields or particles, or any combination thereof.

[0141] Those skilled in the art will further appreciate that the various illustrative logical blocks, modules, circuits, and algorithm steps described in connection with the embodiments disclosed herein can be implemented as electronic hardware, computer software, or combinations of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. Skilled artisans can implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

[0142] The various illustrative logical blocks, and circuits described in connection with the embodiments disclosed herein can be implemented or performed with a general purpose processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. The general purpose processor can be a microprocessor, but in the alternative, the processor can be any conventional processor, controller, microcontroller, or state machine. A processor can also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration.

[0143] The steps of a method or algorithm described in connection with the embodiments disclosed herein can be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module can reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor such that the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium can be integral to the processor. The processor and the storage medium can reside in an ASIC. The ASIC can reside in a user terminal. In the alternative, the processor and the storage medium can reside as discrete components in a user terminal.

[0144] In one or more exemplary embodiments, the functions described can be implemented in hardware, software, firmware, or any combination thereof. If implemented in software as a computer program product, the functions can be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media can be any available media that can be accessed by a computer. By way of example, and not limitation, such computer-readable media can comprise RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. Also, any connection is properly termed a computer-readable medium. For example, if the software is transmitted from a website, server, or other remote source using a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of medium. Disk and disc, as used herein, include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk and Blu-ray® disc where disks usually reproduce data magnetically, while discs reproduce data optically with lasers. Combinations of the above should also be included within the scope of computer-readable media.

[0145] The above-described embodiments are provided as illustrative examples of the application and are not intended to limit the present application. Numerous modifications and adaptations will be apparent to those skilled in the art without departing from the spirit and scope of the present application, as defined in the following claims.​

Claims

1. A method for online monitoring and early warning of the air conditioning status of subway vehicles, characterized in that, include: S1) The host obtains the air conditioner status data and sends it to the first monitoring device. The air conditioner status data is obtained by monitoring the air conditioner control board through the status online monitoring node. The air conditioner status data includes the unit's working status, the unit's power supply status, the unit's communication status, the air conditioner's return air temperature, the air conditioner's fresh air temperature, the air conditioner's set temperature, and the air conditioner's service time. S2) The intelligent operation and maintenance platform obtains the air conditioner status data as the input dataset and uses the pre-trained air conditioner control board remaining life prediction model to generate the remaining predicted life of the air conditioner control board. S3) The intelligent operation and maintenance platform generates early warning information based on the remaining predicted lifespan of the air conditioning control board and sends it to the second monitoring device.

2. The method for online monitoring and early warning of the air conditioning status of subway vehicles according to claim 1, characterized in that, The unit's operating status includes the status of the overall functional area of ​​the control board, the functional area of ​​the central control board, the status of the power supply functional area, the status of the communication functional area, the status of the temperature acquisition functional area, the status of the output control functional area, and the status detection functional area.

3. The method for online monitoring and early warning of the air conditioning status of subway vehicles according to claim 2, characterized in that, The control model for the overall functional area state of the control board includes: F 控制板 =F 中央控制 +F 供电 +F 通讯 +F 温度采集 +F 输出控制 +F 状态监测 ; F 中央控制 =f 中央控制 (x1,x2,...,x n ); F 供电 =f 供电 (y1,y2,...,y n ); F 通讯 =f 通讯 (z1,z2,...,z n ); F 温度采集 =f 温度采集 (m1,m2,...,m n ); F 输出控制 =f 输出控制 (n1,n2,...,n n ); F 状态检测 =f 状态检测 (p1,p2,...,p n ); Among them, F 控制板 This refers to the overall functional area status of the control board; F 中央控制 The states of the functional areas of the central control panel are x1, x2, ..., x. n The status of the components contained within the functional area of ​​the central control board; F 供电 The states of the power supply functional areas are y1, y2, ..., y n The status of components contained within the power supply functional area; F 通讯 The states of the communication functional areas are z1, z2, ..., z. n The status of components contained within the communication functional area; F 温度采集 The states of the temperature acquisition functional areas are m1, m2, ..., m n The status of the components included in the temperature acquisition functional area; F 输出控制 The output control function area states are n1, n2, ..., n n To control the status of the components contained within the output control function area; F 状态监测 For the state detection function area state, p1, p2, ..., p n This refers to the status of the components contained within the status detection function area.

4. The method for online monitoring and early warning of the air conditioning status of subway vehicles according to claim 1, characterized in that, The remaining life prediction model for the air conditioning control board uses a CNN-LSTM prediction model, and the pre-training method for the remaining life prediction model of the air conditioning control board includes: The lifespan data of the air conditioning control board is obtained based on the performance parameters of the key components of the air conditioning control board and related maintenance records. The air conditioner status data and the air conditioner control board lifespan data are used as inputs and normalized to be divided into training datasets and test datasets. The air conditioner control panel remaining life prediction model is trained using the training dataset and the test dataset, and corrected using performance evaluation to obtain the pre-trained air conditioner control panel remaining life prediction model.

5. The method for online monitoring and early warning of the air conditioning status of subway vehicles according to claim 4, characterized in that, The normalization calculation expression is as follows: Among them, X * This represents the normalized data; X represents the input dataset; X max This represents the maximum value in the input dataset; X min This represents the minimum value in the input dataset.

6. The method for online monitoring and early warning of the air conditioning status of subway vehicles according to claim 4, characterized in that, The performance evaluation uses the root mean square error (RMSE) and the coefficient of determination (R²). 2 The evaluation and calculation expressions include: Where N represents the total number of samples in the test dataset; This represents the predicted remaining lifetime of the data sample in the i-th test dataset; This represents the true value of the data sample in the i-th test dataset; RUL mean This represents the average of the actual remaining lifespan of the total number of samples in the test dataset.

7. The method for online monitoring and early warning of the air conditioning status of subway vehicles according to claim 1, characterized in that, Step S3) further includes: When the remaining predicted lifespan of the air conditioning control board is greater than or equal to 16 years but less than 20 years, it is judged as low risk and the warning information is not sent. When the remaining predicted lifespan of the air conditioning control board is greater than or equal to 9 years but less than 16 years, it is judged to be of medium risk, and the warning information is sent to the first monitoring device. When the remaining predicted lifespan of the air conditioning control board is greater than or equal to 0 but less than 9 years, it is judged as high risk, and the warning information is sent to the monitoring equipment.

8. The method for online monitoring and early warning of the air conditioning status of subway vehicles according to claim 2, characterized in that, The parameter settings of the CNN-LSTM prediction model include the parameter settings of the CNN network model and the parameter settings of the LSTM network model. The parameter settings of the CNN network model include the CNN activation function, network convolutional layers and pooling operations. The parameter settings of the LSTM network model include the LSTM network layers, LSTM activation function, loss function and optimizer settings.

9. The method for online monitoring and early warning of the air conditioning status of subway vehicles according to claim 8, characterized in that, The CNN uses ReLU activation function, the network has two convolutional layers, the pooling operation is max pooling, the LSTM network has two layers, the LSTM activation function is Tanh activation function, the loss function is mean squared error loss function, and the optimizer is ADMA optimizer.

10. A subway vehicle air conditioning status online monitoring and early warning system, using the subway vehicle air conditioning status online monitoring and early warning method as described in claims 1-9, characterized in that, include: The status online monitoring node is used to acquire air conditioning status data, which includes unit operating status, unit power supply status, unit communication status, air conditioning return air temperature, air conditioning fresh air temperature, air conditioning set temperature, and air conditioning service time. The intelligent operation and maintenance platform is used to take the air conditioner status data as input and use a pre-trained air conditioner control board remaining life prediction model to generate the remaining predicted life of the air conditioner control board. The intelligent operation and maintenance platform is used to generate early warning information based on the remaining predicted life of the air conditioner control board and send the early warning information to the second monitoring device. The host is used to receive the air conditioner status data obtained by the online status monitoring node and transmit the air conditioner status data to the intelligent operation and maintenance platform through the bus. The host is also used to send the air conditioner status data to the first monitoring device.

Images