A smart carbon reduction analysis and control method and system for a city railway

Abstract

This invention relates to a method and system for intelligent carbon reduction analysis and control of urban rail transit. The method includes: an intelligent server performing intelligent data analysis to generate local control data; the urban rail transit system using this local control data to perform intelligent control of its equipment, collect data from the equipment, and send the real-time collected data to the intelligent server; and sending the local control data to the urban rail transit equipment when a push timing occurs; this push timing is a time of reliable communication, timing, and storage. This invention fully utilizes the development opportunities brought about by artificial intelligence and big data technologies and the decrease in hardware resource deployment costs. By leveraging the computing power of the server, it acquires diverse intelligent control data applicable to the current driving scenario, thereby meeting the intelligent driving needs for multiple objectives such as carbon reduction and fulfilling the requirements of intelligent transportation.

Description

[Technical Field]

[0001] This invention belongs to the field of intelligent transportation technology, and in particular relates to an intelligent carbon reduction analysis and control method and system for urban railways. [Background Technology]

[0002] As a convenient, clean, efficient, and carbon-reducing mode of transportation, urban rail transit, including suburban railways, is an important component of building green cities and developing green transportation in metropolitan areas, and has experienced rapid development in recent years. As the primary commuting tool for urban residents, it has seen the most rapid development within the entire rail transit system, occupying an extremely important position due to its large passenger flow, high frequency, and wide social impact. Besides its high efficiency, large capacity, comfort, safety, and speed, urban rail transit systems, including suburban railways, also involve huge investments and long construction periods. Therefore, the construction and operation of urban rail transit will have a series of impacts on the surrounding environment, and the resulting environmental problems cannot be ignored. On the one hand, subway construction requires a large amount of building materials, generating a large amount of solid waste, dust, noise, and environmental vibration. On the other hand, subway operation consumes a large amount of electricity and produces negative environmental effects such as electromagnetic radiation and environmental vibration. With the continuous increase in urban transportation demand, these problems will continue to worsen. Therefore, higher requirements are placed on rail transit. Safe operation is a basic requirement, and on this basis, intelligent operation is needed to meet multiple objectives, which is in line with current development characteristics.

[0003] Intelligent transportation, as a crucial development direction for modern urban traffic management, deeply integrates big data and artificial intelligence technologies. It aims to achieve efficient, safe, and sustainable development of transportation systems through data-driven and intelligent decision-making. Big data technology provides the system with holistic perception capabilities by collecting and processing massive amounts of traffic data in real time, such as vehicle trajectories, traffic light status, and environmental information. Artificial intelligence technology, relying on machine learning and deep learning algorithms, mines patterns from the data and generates optimization strategies. The synergistic application of these technologies significantly improves the accuracy and adaptability of multi-objective intelligent control.

[0004] In intelligent transportation systems, multi-objective optimization is one of the core challenges. Traffic control can be optimized for multiple objectives, such as efficiency and carbon emissions. Big data technology, through historical and real-time data analysis, can quantify the correlations and conflicts between different objectives, such as the trade-off between time and energy consumption, and the balance between acceleration and comfort. Artificial intelligence technology, through reinforcement learning and multi-objective optimization algorithms, dynamically adjusts control strategies to achieve the optimal solution for the combination of objectives while ensuring safety. For example, predicting traffic flow using deep learning models can optimize traffic light timing to reduce congestion; combining vehicle energy consumption models can plan optimal driving curves to reduce carbon emissions; and intelligent cruise control based on real-time traffic conditions can coordinate the relationship between speed, acceleration, and energy consumption, avoiding additional losses caused by sudden acceleration / deceleration. Furthermore, the integration of big data and artificial intelligence also supports the dynamic adaptability of transportation systems. Through edge computing and cloud collaboration, the system can respond to emergencies in real time and quickly adjust control strategies. The ability to collaboratively optimize multiple objectives not only improves traffic efficiency but also provides key technological support for the low-carbon and intelligent transformation of smart cities.

[0005] In recent years, the rapid development of artificial intelligence (AI) technology and the continuous optimization of hardware resources have brought unprecedented development opportunities to intelligent transportation systems. On the one hand, with advancements in chip manufacturing processes and the widespread adoption of dedicated AI accelerator chips, the cost of computing hardware has been continuously decreasing, while computing power has increased exponentially. This makes it possible to deploy high-performance AI models on a large scale, especially enabling deep learning-based traffic prediction, optimization, and decision-making systems to run efficiently on edge devices and cloud servers. On the other hand, the maturity of cloud computing and distributed computing technologies has made real-time processing and analysis of massive amounts of traffic data a reality, providing strong computing support for intelligent transportation systems. These technological advancements have had a profound impact on the development prospects of intelligent transportation. First, powerful computing capabilities have significantly improved the training and inference speed of complex traffic models. For example, reinforcement learning-based dynamic path planning algorithms can complete decisions in milliseconds, significantly improving the response speed of traffic systems. Second, the widespread availability of hardware resources has lowered the deployment threshold for intelligent transportation systems, enabling urban rail transit to handle more control data. More importantly, servers with high computing power can support the fusion analysis of multimodal data, such as simultaneously processing video surveillance, radar detection, vehicle-mounted terminal, and meteorological data, thereby providing a more comprehensive data foundation for multi-objective intelligent control. In the future, with the further maturation of 5G / 6G communication technology and vehicle-road cooperative systems, combined with high-performance computing resources, intelligent transportation systems will achieve more precise real-time control, more efficient energy utilization, and a safer travel experience, ultimately driving urban transportation towards comprehensive intelligence, greening, and humanization. How to fully utilize the development opportunities brought about by artificial intelligence and big data technologies, as well as the reduction in hardware resource deployment costs, is a technical problem to be solved. Based on the above problems, this invention can fully utilize the computing power of servers to acquire diverse intelligent control data applicable to the current driving scenario and transmit it to the urban rail transit device at a reliable time. The urban rail transit device itself does not need strong computing and reasoning capabilities; it only requires storage resources to smoothly acquire sufficient intelligent control data for intelligent control guidance and intelligent driving, thereby meeting multi-objective intelligent driving needs and satisfying the requirements of intelligent transportation. [Summary of the Invention]

[0006] To address the aforementioned problems in the prior art, this invention proposes an intelligent carbon reduction analysis and control method and system for urban rail transit, the system comprising: an intelligent server and multiple urban rail transit devices;

[0007] Intelligent servers and urban rail transit devices; urban rail transit may be various forms of rail transportation within a city, such as trams and subways; and urban rail transit devices are used for transportation and / or transportation control within a city.

[0008] The intelligent server performs intelligent data analysis to generate local control data; specifically: it determines the dynamic weight of each type of data acquisition based on recently acquired data; it determines the local input data set based on the dynamic weight and the benchmark data acquisition; it inputs each local input data in the local input data set into the control model to obtain the corresponding local control data; and it sends the local control data to the urban rail transit device when the push timing arrives; the push timing is a reliable time for communication, time and / or storage.

[0009] Specifically: the instability of each type of collected data within the most recent time period is determined; based on the instability, the dynamic weight of each type of collected data is determined; the range and number of changes of the baseline collected data for that type are set so that the larger the dynamic weight value, the larger the range and the more changes of the data for that type; the baseline collected data is changed within the range of change based on the baseline collected data to obtain local input data; the local input data after the number of changes constitutes the local input data set.

[0010] The urban rail transit device performs intelligent control based on the local control data, collects data from the urban rail transit device, and sends the collected data to the intelligent server in real time.

[0011] There are multiple control models, each corresponding to an intelligent control objective. Based on the intelligent control objective, a corresponding control model is selected. For each selected control model, each set of local input data is input into the control model to obtain local control data corresponding to each set of local input data, forming a local control data group. When there are multiple control objectives, multiple sets of local control data corresponding to each intelligent control objective are obtained. The local control data groups are merged to obtain local control data.

[0012] Furthermore, the step of sending the collected data to the intelligent server specifically involves: when the trolley is in operation, collecting driving data and environmental data in real time and sending them to the intelligent server.

[0013] Furthermore, the push timing is when one or more of the following conditions are met: the urban rail device reaches the destination location, the communication quality is good, the difference between the local control data stored locally by the urban rail device and the local control data to be pushed by the intelligent server is large, and the urban rail device has sufficient storage space.

[0014] Furthermore, the most recent time length T includes T unit time intervals t, t∈1~T; determining the instability of each type of collected data within the most recent time length, and determining the dynamic weight of the collected data based on the instability; specifically includes the following steps:

[0015] Step SE1: Acquire k-type collected data within time interval t;

[0016] Step SE2: Determine the unstable entropy of the k-type collected data; the smaller the unstable entropy, the more unstable the data, and vice versa, the closer the entropy is to 1, the more stable the data.

[0017] Step SE3: Determine the dynamic weights of type k based on the unstable entropy.

[0018] Furthermore, the instability of the collected data is determined. When the instability is high, local input data is generated based on a genetic algorithm to supplement the local input data set. Specifically, this includes the following steps:

[0019] Step SV1: Determine the instability level of the collected data; if the instability level is high, proceed to the next step; otherwise, end.

[0020] Step SV2: Generate crossover and mutation data based on a genetic algorithm, and use it to supplement the local input dataset; specifically, it includes the following steps:

[0021] Step SV21: Perform a mutation operation on the k-type collected data in the first collected data to obtain the second collected data; where: the first collected data is the current collected data;

[0022] Step SV22: Perform cross-processing on the first and second collected data to obtain the third collected data;

[0023] Step SV23: Supplement the local input data set with the third acquired data.

[0024] Furthermore, the process of merging local control data groups to obtain local control data specifically involves: by default, sending the union of local control data groups as the final determined local control data to the municipal railway device; when the push resources of the municipal railway device are limited, or when the push time, storage space of the municipal railway device, and / or poor communication quality lead to limited push resources, the union of local control data groups is sent as the final determined local control data to the municipal railway device.

[0025] A method for intelligent carbon reduction analysis and control of urban rail transit, the method being based on the aforementioned intelligent carbon reduction analysis and control system for urban rail transit, comprising:

[0026] Step S1: The intelligent server performs intelligent data analysis to generate local control data; specifically: determining the dynamic weight of each type of collected data based on recently collected data; and determining the local input data set based on the dynamic weight and the benchmark collected data.

[0027] Step S2: The urban rail device is used for intelligent control of the urban rail device based on the local control data, to collect data of the urban rail device, and to send the real-time collected data to the intelligent server; and to send the local control data to the urban rail device when the push timing arrives; the push timing is a time when communication, time and storage are reliable.

[0028] A smart carbon reduction analysis and control platform for urban rail transit includes the aforementioned smart carbon reduction analysis and control system for urban rail transit.

[0029] A computer cloud computing server includes the aforementioned intelligent carbon reduction analysis and control system for urban rail transit.

[0030] An intelligent server, including the aforementioned intelligent carbon reduction analysis and control system for urban rail transit.

[0031] The beneficial effects of this invention include:

[0032] (1) It can make full use of the computing power of the intelligent server to obtain diverse intelligent control data that can be used in the current driving scenario, without affecting the reliable timing of normal driving to transmit to the urban rail device; while the urban rail device itself does not need to have strong computing and reasoning capabilities, it only needs the support of storage resources to smoothly obtain sufficient intelligent control data for intelligent control guidance of intelligent driving, thereby meeting the multi-objective intelligent driving needs, meeting the needs of intelligent transportation, and meeting the high-quality pursuit of low carbon, high efficiency and high speed on the basis of normal driving.

[0033] (2) With the support of the computing power of intelligent servers, it is suitable for multi-objective intelligent driving pursuit and has strong compatibility; furthermore, it can detect abnormal situations such as unstable data collection in real time through unstable entropy, and can effectively supplement abnormal situations through genetic mutation, thereby providing effective support for the derivation of diverse intelligent control data, and improving the level of intelligence on the basis of normal driving. [Attached Image Description]

[0034] The accompanying drawings, which are provided to further illustrate the invention and form part of this application, are not intended to unduly limit the invention. In the drawings:

[0035] Figure 1 A schematic diagram of the intelligent carbon reduction analysis and control method for urban railways provided by the present invention.

Detailed Implementation Methods

[0036] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. The illustrative embodiments and descriptions are only used to explain the present invention and are not intended to limit the present invention.

[0037] This invention proposes an intelligent carbon reduction analysis and control system for urban rail transit; the system includes: urban rail transit equipment and an intelligent server;

[0038] The intelligent server is used to perform intelligent data analysis to generate local control data; specifically: determining the dynamic weight of each type of acquired data based on recently acquired data; determining the local input data set based on the dynamic weight and the benchmark acquired data; inputting each input control model in the local input data set to obtain the corresponding local control data; the local control data is a set containing multiple intelligent control strategies corresponding to each local input data, including a variety of intelligent control strategies;

[0039] The urban rail device is used for intelligent control of the urban rail device based on the local control data, to collect data from the urban rail device, and to send the real-time collected data to the intelligent server; the collected data is one or more types k, k∈1~K; where: K is the number of types of collected data;

[0040] The step of sending the collected data to the intelligent server specifically involves: collecting driving data and environmental data in real time and sending them to the intelligent server when the trolley is in operation.

[0041] Preferably, the intelligent server is a big data server; the big data server connects to multiple urban rail transit devices; it is used to collect the collected data from the multiple urban rail transit devices, construct historical collected data based on the collected data; and perform intelligent analysis and control on the multiple urban rail transit devices; the amount of real-time collected data is relatively small and will not impose a burden on the current journey in terms of computing, communication and storage; nor does it require excessive software and hardware support for transmission reliability.

[0042] Preferably, the intelligent server is an intelligent auxiliary server, which provides local control data to provide intelligent control assistance for the operation of the urban rail transit device;

[0043] Preferably, the historical data collected includes factory test data, etc. The historical data collected is organized based on the attribute information of the urban railway equipment. In the subsequent analysis process, the historical data collected is provided for intelligent analysis based on the attributes of the urban railway equipment.

[0044] Alternative: The intelligent server is a cloud computing server; the cloud computing server connects to multiple urban rail transit devices and performs intelligent analysis and control on the multiple urban rail transit devices; when a trolleybus is connected, trolleybus attribute data is acquired, and urban rail transit devices are analyzed and controlled based on the attribute data; when attribute information changes, the changed attribute data needs to be collected and updated.

[0045] Preferred configuration: After the tram starts the control system or intelligent control process, it enters an active state. Based on this active state, the intelligent server is triggered to start intelligent control. The intelligent server receives the collected data in real time, performs intelligent data analysis to generate local control data, and sends the local control data to the urban rail device when the push timing arrives. This push timing is a time when communication, time, and storage are reliable.

[0046] Preferably, the push timing is when one or more of the following conditions are met: the urban rail device reaches the destination location, communication quality is good, there is a large difference between the local control data stored locally by the urban rail device and the local control data to be pushed by the intelligent server, and the urban rail device has sufficient storage space; furthermore, the intelligent server stores push history information (or fingerprint information of push history information), and the difference between the local control data stored locally by the urban rail device and the local control data to be pushed by the intelligent server is determined by comparing with the push history information, thereby determining whether the push timing has been reached;

[0047] Preferably, the collected data includes driving data, environmental data, and / or attribute data; wherein: the driving data includes position, speed, power description, traction transmission performance, acceleration, deceleration, output power, etc.; driving data is acquired through sensing devices installed on the tram, such as speed sensors, position sensors, or acceleration sensors; environmental data includes driving gradient, temperature, weather, road conditions, driving scenario, etc.; for example, acquired through tilt sensors installed on the vehicle to measure slope or descent, or friction sensors installed on the track to measure humidity, icing, or snow, or 2D or 3D image data of the driving scenario acquired through image sensors; the attribute data is data related to the urban rail transit device itself and its driving plan, and this data is relatively static; attribute data includes: tram type, identification, driving plan, crew members, load, etc.

[0048] Preferred configuration: When the intelligent server receives attribute data, it creates a tram control data block for the tram; when the tram is in an active state, it creates or activates an idle control process to load the tram control block and performs intelligent analysis and control based on real-time collected data;

[0049] The dynamic weight of each type of collected data is determined based on recently collected data. Specifically, the instability of each type of collected data within the most recent time period T is determined, and the dynamic weight of the collected data is determined based on the instability. Wherein, the most recent time period T contains T unit time intervals t, t∈1~T.

[0050] Specifically, the steps include the following:

[0051] Step SE1: Acquire k-type collected data within time interval t

[0052] Preferred method: Obtain the most recent time period T from the tram control data block corresponding to the tram; obviously, when the tram data is incomplete, unstable, or discontinuous, the tram uses its own inherent local control data for tram control; this inherent local control data will not be replaced with the push; the inherent local control data contains the inherent control strategy;

[0053] Step SE2: Determine the unstable entropy uc of the k-type collected data. k Specifically, the unstable entropy uc of type k collected data is calculated based on the following formulas (1)-(3). k The smaller the instability entropy, the more unstable the system; conversely, the closer the entropy is to 1, the more stable the system. Where: It is the normalized value of the k-type collected data over the t time interval;

[0054]

[0055] Step SE3: Determine the dynamic weight w based on the unstable entropy. k Specifically, dynamic weights are set for k-type collected data based on the magnitude of the unstable entropy, so that the larger the unstable entropy, the higher the dynamic weight value; specifically, the following formula (4) is used to set dynamic weights w for k-type collected data. k ;

[0056] w k =1-uc k (4);

[0057] Alternative: Determining dynamic weights w based on stable weights and unstable entropy. k Specifically, this involves obtaining the preset stable weights st_w for k-type collected data. k Set unstable weights w_uc k =1-uc k The dynamic weight w is set based on the following formula (5) or (6). k ;in:

[0058] β1 and β2 are weighting coefficients;

[0059] w k =(β1×st_w k +β2×w_uc k (5);

[0060]

[0061] Preferred configuration: β1 + β2 = 1; for example: β1 = 0.6; β2 = 0.4;

[0062] Preferred: The stable weight is a preset value, set according to the inherent importance of the collected data type k to the local control data;

[0063] The process of determining the local input data set based on the dynamic weight and benchmark data involves: setting the variation range (and number of changes) of the benchmark data based on the dynamic weight, such that a larger dynamic weight value results in a larger variation range and more changes; changing the benchmark data within the variation range based on the benchmark data to obtain local input data, and the local input data after completing the required number of changes constitutes the local input data set; specifically, it includes the following steps:

[0064] Step SX1: Acquire baseline data (bcd) k ), k = 1 ~ K (that is, (bcd1, bcd) # ,…,bcd > The baseline data is a set of key data. Specifically, the current data, the empirical data set for the current environment, or the data or predicted data determined by the driving plan and corresponding to the current driving position are used as the baseline data. In other words, the current data is the data that reflects or conforms to the current trolley status and scenario as much as possible.

[0065] Step SX2: Arrange the dynamic weights in ascending order to form a dynamic weight sequence. <sw k >;At this time sw k The dynamic weights corresponding to the sorted k-type collected data;

[0066] Step SX3: Acquire the k-th type of benchmark acquisition data bcd k Set the initial value of k to 1.

[0067] Step SX4: Based on dynamic weights sw k Set the range of variation for k-type baseline acquisition data [lcd] k hcd k ]; with dynamic weights sw k To adjust the scale, the data is adjusted upwards and / or downwards based on the baseline data to form a range of variation; specifically, the range of variation is determined using the following formulas (7) and (8); where: min(cd k ) and max(cd k These are the minimum and maximum values ​​of the data collected for type k, respectively.

[0068] LCD k =bcd k -sw k ×(bcd k -min(cd k ))(7);

[0069] hcd k =bcd k +sw k ×(max(cd k )-bcd k (8);

[0070] Step SX5: Set the number of changes N k ∝sw k This step is optional.

[0071] Step SX6: Set k = k + 1; if k > K, then end; otherwise, return to step SX3; by determining the range of change in ascending order, the subsequent data types can be continuously amplified under the constraint of the data types collected within the smaller range of change.

[0072] The process involves varying the baseline acquisition data within a certain range based on the baseline acquisition data to obtain local input data. Specifically, this involves extracting acquisition data that meets the inclusion criteria from historical acquisition data and placing it into the local input data set. The inclusion criteria are that the acquisition data of each type k are within its variation range [lcd]. k hcd k The data collected within the set is combined such that, in this local input data set, and / or the number of different values ​​of each data type k in this set occurs equal to (close to) the number of changes N. k Historical data collection comes from current municipal railway equipment, other municipal railway equipment, test data, expert system data, etc.; that is to say, in historical experience, the data collection data of each type k satisfies the combination of data collection data within the range of variation.

[0073] Preferably, the placement conditions also include that the placed data is typical data; typicality can be determined by the number of times it appears in historical data; the placed data is made to conform to the numerical constraints between different data types in historical operation patterns by combining the data; in addition, the distribution of values ​​is as dispersed as possible to cover more control scenarios and reflect random characteristics.

[0074] Alternative: The step of varying the benchmark data within a range based on the benchmark data to obtain local input data specifically includes the following steps:

[0075] Step SU1: For type k, obtain the range of variation [lcd] from historical data. k hcd k N within ] k N are the collected data values ​​corresponding to type k, and the Nk Each collected data value constitutes a subset of k types of numerical values; repeat this step until all types k have been processed, and then you will get K subsets of k types of numerical values.

[0076] Step SU2: Randomly select one data value from each k-type numerical subset to form local input data of the K-ary and put it into the local input data set; repeat this step until the local input data set reaches the preset size NL; of course, you can judge whether the local input data of the K-ary satisfy the constraint relationship before putting it in; when selecting one data value, select as dispersedly as possible to cover each element in each k-type data subset;

[0077] Replaceable: Determine the instability level of the collected data. When the instability level is high, generate local input data based on a genetic algorithm to supplement the local input data set; specifically, this includes the following steps:

[0078] Step SV1: Determine the instability level uL of the collected data; specifically: when (when If the value is greater than or equal to the cutoff threshold, the instability level is determined to be high, and the process proceeds to the next step; otherwise, the process ends. However, in some cases, the instability level determination obtained by judging only the most recent time length T may fluctuate. In such cases, the following alternative method can be used; where: It's sw k The mean; It is w k The mean;

[0079] Preferably, the cutoff threshold is a preset value, for example, 0.1 to 0.3;

[0080] Alternatively, step SV1 specifically involves: determining the variance deviation of the unstable entropy over multiple time lengths T, using this deviation to indicate the degree of instability; further, calculating the degree of instability based on the following equations (9)-(11); when the degree of instability uL is greater than the instability threshold, proceeding to the next step; otherwise, ending; wherein: The unstable entropy of type k data collected in the most recent ht-th time length; ht∈1~HT, HT is the number of time lengths; each time length ht contains T unit time intervals;

[0081]

[0082] Preferably, the instability threshold is a preset value, for example, 0.05 to 0.2; HT time lengths are the most recent HT time lengths; HT includes the HT consecutive most recent time lengths T before the current time; that is, the instability entropy is calculated based on the most recent historical data.

[0083] Step SV2: Generate crossover and mutation data based on a genetic algorithm, and use it to supplement the local input dataset; specifically, it includes the following steps:

[0084] Step SV21: For the first collected data (d1) k Data d1 of type k in ) k Perform mutation operation to obtain the second collected data (d2) k ); where: the first collected data is the baseline collected data, the collected data at any time interval t, and one or more of the currently collected data; specifically: the second collected data is calculated using the following formulas (12) and (13); where: γ k It is the coefficient of variation, and random[-1,1] is a random number between [-1,1].

[0085] d2 k =d1 k ×(1+γ k ×random[-1,1]) (12);

[0086] d2 k ∈[min(cd k ),max(cd k )](13);

[0087] Preferred: Based on SW k Set γ k For example: setting γ k ∝sw k ;

[0088] Step SV22: Perform a cross operation on the first and second acquired data to obtain the third acquired data (d3). k Specifically, the third collection data (d3) of variation is calculated using the following formula (14). k ); where random[0,1] is a random number between [0,1];

[0089] d3 k =d1 k ×random[0,1]+d2 k ×(1-random[0,1]) (14);

[0090] Step SV23: Supplement the local input data set with the third acquired data;

[0091] The step involves inputting each local input data in the local input data set into the control model to obtain the corresponding local control data; specifically, the control model is an expert system, and each local input data is input into the decision control model to obtain the local control data corresponding to each local input data.

[0092] Preferably, there are multiple control models, each control objective is for a specific intelligent control objective. First, a corresponding control model is selected based on the intelligent control objective. For each control model, each local input data is input into the decision control model to obtain local control data corresponding to each local input data. When there are multiple control objectives, multiple sets of local control data corresponding to one intelligent control objective are obtained.

[0093] Furthermore: the control model is a pre-trained artificial intelligence model. The input of the artificial intelligence model is the collected data, and the output is decision control data as local control data. Specifically: the current intelligent control target is obtained, the artificial intelligence model corresponding to the current intelligent control target is obtained, and each local input data is input into the artificial intelligence model to obtain the decision control data corresponding to each local input data.

[0094] Preferably, the artificial intelligence model is a neural network model, a feedback neural network model, or a deep neural network model;

[0095] Preferably, the intelligent control target is an independent indicator, such as one or more of the following: time, low carbon emissions, acceleration, deceleration, speed, energy consumption, and loss.

[0096] Preferably, the intelligent control objective is a comprehensive indicator, such as safety, comfort, reliability, and economy. When the control objective is a comprehensive indicator, it is decomposed into corresponding independent indicators. An input control model is selected for each independent indicator, and each local input data in the local input data set is sequentially input into the selected input control model to obtain local control data corresponding to each selected input control model. The collection of local control data is used as the final determined local control data and sent to the urban rail transit device.

[0097] Preferred method: When the push resources of the urban rail transit device are limited, the union of local control data is used as the final determined local control data and sent to the urban rail transit device; when the push resources are limited due to factors such as push time, storage space of the urban rail transit device, and poor communication quality, the size of the push data can be appropriately limited by the union method. Of course, other methods of limiting the size of the set can also be used, which will not be elaborated here.

[0098] As attached Figure 1 As shown, based on the same inventive concept, the present invention also provides a method for intelligent carbon reduction analysis and control of urban railways, the method being used for intelligent control of urban railways based on the intelligent carbon reduction analysis and control system of urban railways;

[0099] The method includes:

[0100] Step S1: The intelligent server performs intelligent data analysis to generate local control data; specifically: determining the dynamic weight of each type of collected data based on recently collected data; and determining the local input data set based on the dynamic weight and the benchmark collected data.

[0101] Step S2: The urban rail device is used for intelligent control of the urban rail device based on the local control data, to collect data of the urban rail device, and to send the real-time collected data to the intelligent server; and to send the local control data to the urban rail device when the push timing arrives; the push timing is a time when communication, time, and storage are reliable;

[0102] Based on the same inventive concept, the present invention also provides an intelligent carbon reduction analysis and control server for urban railways, the server comprising the intelligent carbon reduction analysis and control system for urban railways;

[0103] Based on the same inventive concept, the present invention also provides an intelligent carbon reduction analysis and control platform for urban railways, the platform including the intelligent carbon reduction analysis and control system for urban railways;

[0104] A computer program (also referred to as a program, software, software application, script, or code) can be written in any form of programming language, including assembly or interpreted languages, declarative or procedural languages, and can be deployed in any form, including as a standalone program or as a module, component, subroutine, object, or other unit suitable for use in a computing environment. A computer program may, but does not necessarily, correspond to a file in a file system. A program can be stored as part of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single file dedicated to said program, or in multiple co-located files (e.g., a file storing one or more modules, subroutines, or code portions). A computer program can be deployed to execute on a single computer or on multiple computers located at a single site or distributed across multiple sites and interconnected by a communications network.

[0105] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0106] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0107] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0108] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A smart carbon reduction analysis and control system for urban railways, characterized in that, Includes intelligent servers and urban rail transit equipment; The intelligent server performs intelligent data analysis to generate local control data; specifically: it determines the dynamic weight of each type of data acquisition based on recently acquired data; it determines the local input data set based on the dynamic weight and the benchmark data acquisition; it inputs each local input data in the local input data set into the control model to obtain the corresponding local control data; the local control data is a set containing multiple intelligent control strategies corresponding to each local input data, including a variety of intelligent control strategies. The step of determining the dynamic weight of each type of collected data based on recently collected data specifically involves: determining the instability of each type of collected data within the most recent time period T, and determining the dynamic weight of the collected data based on this instability; the most recent time period T contains T unit time intervals t, t∈1~T; the determination of the instability of each type of collected data within the most recent time period and the determination of the dynamic weight of the collected data based on this instability specifically includes the following steps: Step SE1: Acquire k-type collected data within time interval t; Step SE2: Determine the unstable entropy of the k-type collected data; the smaller the unstable entropy, the more unstable the data, and vice versa, the closer the entropy is to 1, the more stable the data. Step SE3: Determine the dynamic weights of type k based on the unstable entropy; The process of determining a local input data set based on the dynamic weight and benchmark data involves: setting the range and number of changes of the benchmark data based on the dynamic weight, such that the larger the dynamic weight value, the larger the range and the more changes; and, based on the benchmark data, changing the benchmark data within the range to obtain local input data. The local input data after completing the number of changes constitutes the local input data set. The baseline data is a key data set; the current data, the empirical data set for the current environment, or the data set or predicted data corresponding to the current driving position determined by the driving plan are used as the baseline data. When the push timing arrives, local control data is sent to the urban rail device; The push timing is when one or more of the following conditions are met: the urban rail device reaches the destination location, the communication quality is good, the difference between the local control data stored locally by the urban rail device and the local control data to be pushed by the intelligent server is large, and the urban rail device has sufficient storage space. The urban rail system uses this local control data for intelligent control of the urban rail system; it is also used for data acquisition of the urban rail system and sends the acquired data to the intelligent server in real time. There are multiple control models, each corresponding to an intelligent control objective. A corresponding control model is selected based on the intelligent control objective. For each selected control model, each set of local input data is input into the control model to obtain local control data corresponding to each set of local input data, forming a local control data group. When there are multiple control objectives, multiple sets of local control data corresponding to each intelligent control objective are obtained. The local control data groups are merged to obtain local control data. The control model is a pre-trained artificial intelligence model.

2. The intelligent carbon reduction analysis and control system for urban railways according to claim 1, characterized in that, The collected data is sent to the intelligent server; specifically, when the trolley is in operation, driving data and environmental data are collected in real time and sent to the intelligent server.

3. The intelligent carbon reduction analysis and control system for urban railways according to claim 2, characterized in that, The instability level of the collected data is determined. When the instability level is high, local input data is generated based on a genetic algorithm to supplement the local input data set. Specifically, the steps include: Step SV1: Determine the instability level of the collected data; if the instability level is high, proceed to the next step; otherwise, end. Step SV2: Generate crossover and mutation data based on a genetic algorithm, and use it to supplement the local input dataset; specifically, it includes the following steps: Step SV21: Perform a mutation operation on the k-type collected data in the first collected data to obtain the second collected data; where: the first collected data is the current collected data; Step SV22: Perform cross-processing on the first and second collected data to obtain the third collected data; Step SV23: Supplement the local input data set with the third acquired data.

4. The intelligent carbon reduction analysis and control system for urban railways according to claim 3, characterized in that, The process of merging local control data groups to obtain local control data specifically involves: by default, sending the union of local control data groups as the final determined local control data to the municipal railway device; when the push resources of the municipal railway device are limited, or when the push time, storage space of the municipal railway device, and / or poor communication quality lead to limited push resources, the union of local control data groups is sent as the final determined local control data to the municipal railway device.

5. A method for intelligent carbon reduction analysis and control of urban railways, characterized in that, The method is based on the intelligent carbon reduction analysis and control system for urban railways according to any one of claims 1-4, and includes: Step S1: The intelligent server performs intelligent data analysis to generate local control data; specifically: determining the dynamic weight of each type of collected data based on recently collected data; and determining the local input data set based on the dynamic weight and the benchmark collected data. Step S2: The urban rail device is used for intelligent control of the urban rail device based on the local control data, to collect data of the urban rail device, to send the real-time collected data to the intelligent server, and to send the local control data to the urban rail device when the push time arrives.

6. A smart carbon reduction analysis and control platform for urban railways, characterized in that, Including the intelligent carbon reduction analysis and control system for urban railways as described in any one of claims 1-4.

7. A computer cloud computing server, characterized in that, The cloud computing server is configured to include the urban rail intelligent carbon reduction analysis and control system as described in any one of claims 1-4.

8. An intelligent server, characterized in that, The intelligent server is configured to include the urban rail intelligent carbon reduction analysis and control system as described in any one of claims 1-4.

Images