CN117301928B

CN117301928B - Charging early warning method and device for charging pile

Info

Publication number: CN117301928B
Application number: CN202311223938.5A
Authority: CN
Inventors: 苏尚彬; 马正伟; 董银飞
Original assignee: Beiqi Foton Motor Co Ltd; Shandong Vocational College of Industry
Current assignee: Beiqi Foton Motor Co Ltd; Shandong Vocational College of Industry
Filing date: 2023-09-21
Publication date: 2024-07-16
Anticipated expiration: 2043-09-21

Abstract

The invention discloses a charging early warning method and device for a charging pile, and belongs to the technical field of charging early warning of charging piles; carrying out real-time monitoring and accurate temperature measurement on the charging pile based on an infrared thermal imaging technology, an acoustic wave sensor technology and a laser radar technology, carrying out intelligent identification and remote monitoring on the charging pile based on an RFID technology, carrying out safe storage, verification and transmission on data of the charging pile based on a blockchain technology and an intelligent contract technology, carrying out feature extraction, classification and identification on the data of the charging pile based on a deep learning algorithm, carrying out intelligent early warning of abnormal temperature rise, overtemperature or fire disaster according to temperature data and a fault mode, and displaying early warning states and emergency measures through a user terminal or a monitoring center; the fault identification accuracy and efficiency of the charging pile can be effectively improved, the fire risk and loss of the charging pile are reduced, and the charging experience and the safety feeling of a user are improved.

Description

Charging early warning method and device for charging pile

Technical Field

The invention discloses a charging early warning method and device for a charging pile, and belongs to the technical field of charging early warning of charging piles.

Background

With the popularization and development of new energy automobiles, the charging pile is used as an important supporting facility of the new energy automobiles, and is also increasingly and widely applied. However, various faults or abnormal conditions, such as overload, short circuit, electric leakage, fire and the like, can be encountered in the using process of the charging pile, so that serious threat is brought to the safety and property of the user. Therefore, how to effectively monitor and pre-warn the charging pile and timely discover and process faults or abnormal conditions is a technical problem to be solved in the present.

The existing charging pile monitoring and early warning technology mainly comprises the following steps:

One is based on traditional temperature sensor or smoke transducer to charge the stake and monitor to carry out early warning suggestion through alarm or display. For example, chinese patent publication No. CN212750040U discloses a detection device for smog warning suitable for electric motor car charges, including case, alarm lamp and temperature sensor, case a lateral wall middle part is provided with smoke sensor, smoke sensor one side is provided with the alarm, smoke sensor keeps away from alarm one side is provided with the alarm lamp, alarm lamp one side be provided with the connecting wire of case grafting, the connecting wire is kept away from case one end is provided with temperature sensor. The disadvantages of this technique are: the temperature sensor or the smoke sensor can only monitor a local area and cannot cover the whole charging pile; the alarm or the display can only provide simple early warning signals or information, and cannot provide detailed fault reasons and emergency measures; in addition, this technique also cannot effectively monitor and pre-warn the remote or unattended charging pile.

And the other is to monitor the charging pile based on a video monitoring system and perform data analysis and early warning prompt through an artificial intelligence technology. For example, chinese patent publication No. CN108583317a discloses a charging pile monitoring system, including charging pile, charging pile includes charging module, master controller, video monitor module, input monitor module and output monitor module, charging module's input passes through the relay and connects the electric wire netting, charging module's output connects outside battery charging outfit, input monitor module is used for monitoring the electric power parameter of charging module input, output monitor module is used for monitoring the electric power parameter of charging module output, the electric wire netting, charging module, video monitor module, output monitor module, input monitor module and relay are connected with master controller. The disadvantages of this technique are: the video monitoring system needs a large amount of cameras and network bandwidth, so that high cost and resource waste are caused; the artificial intelligence technology needs a large amount of data training and updating, and has high false alarm rate and false missing rate; in addition, this technique also fails to effectively monitor and pre-warn the internal conditions of the charging stake.

And the charging pile is monitored based on the internet of things technology, and data storage, analysis and early warning prompt are performed through the cloud platform. The disadvantages of this technique are: the technology of the Internet of things requires a large number of sensors and communication modules, so that higher energy consumption and maintenance cost are caused; the cloud platform needs a large amount of servers and storage space, and risks of data security and privacy disclosure exist; in addition, this technique also fails to accurately measure and identify temperature changes and failure modes of the charging stake.

In order to overcome the defects in the prior art, the invention provides a charging early warning method and device for a charging pile.

Disclosure of Invention

The invention aims to provide a charging early warning method and device for a charging pile, which have the advantages of high efficiency, safety, intelligence and convenience, and solve the problems of inaccurate, untimely, detailed, unfriendly and inconvenient fault identification of the traditional charging pile.

In order to achieve the above purpose, the present invention provides the following technical solutions:

The invention relates to a charging early warning method of a charging pile, which comprises the steps of carrying out real-time monitoring, data analysis and early warning prompt on the charging pile, wherein the real-time monitoring comprises the steps of carrying out high-definition monitoring and accurate temperature measurement on the charging pile based on an infrared thermal imaging technology, an acoustic wave sensor technology and a laser radar technology, and carrying out intelligent identification and remote monitoring on the charging pile based on an RFID technology; the data analysis comprises the steps of safely storing, verifying and transmitting the data of the charging pile based on a blockchain technology and an intelligent contract technology, and extracting, classifying and identifying the characteristics of the data of the charging pile based on a deep learning algorithm; the early warning prompt comprises intelligent early warning of abnormal temperature rise, overtemperature or fire disaster according to the temperature data and the fault mode, and early warning states and emergency measures are displayed through a user terminal or a monitoring center.

Preferably, the abnormal temperature rise is identified by a temperature fluctuation score and a temperature convexity score:

The temperature fluctuation score wfs _i is calculated by the following formula:

Wherein t _i,j represents the temperature of the jth sub-region in the ith round of detection, t _i represents the average temperature of all the sub-regions in the ith round of detection, and n1 represents the number of sub-regions. The temperature fluctuation score reflects the degree of change of the temperature in the charging pile area, and the larger the temperature fluctuation score is, the more unstable the temperature is;

The temperature convexity score cfs _i is calculated by the following formula:

wherein t _i,j represents the temperature of the jth sub-zone in the ith round of detection, The average temperature of all subareas in the ith round of detection is represented, and n ₁ represents the number of subareas. The temperature convexity score reflects the peak value or valley value degree of the temperature in the charging pile area, and the larger the temperature is, the less smooth the temperature is;

And determining a temperature detection score of each subregion according to the temperature fluctuation score and the temperature convexity score, and determining the subregion with abnormal temperature according to the temperature detection score.

Preferably, the high-definition monitoring and accurate temperature measurement are carried out on the charging pile based on the infrared thermal imaging technology, and the method comprises the following steps:

(1) The infrared thermal imager emits infrared rays to the charging pile and receives the infrared rays reflected from the charging pile;

(2) The infrared thermal imager converts the received infrared rays into digital signals, calculates temperature information of a target object according to the intensity and the wavelength of the signals, and generates corresponding images;

(3) And displaying an image generated by the infrared thermal imager by the display, and carrying out early warning prompt according to the set threshold value.

Preferably, the method for monitoring the charging pile in real time and measuring the temperature accurately based on the acoustic wave sensor technology comprises the following steps:

(1) The sound wave sensor transmitter transmits sound wave signals to the inside of the charging pile;

(2) The acoustic wave sensor receiver receives acoustic wave signals reflected from the inside of the charging pile and converts the signals into digital signals;

(3) The data processor receives the digital signal, calculates the temperature information of the target object according to the propagation time and frequency of the signal, and generates corresponding data;

(4) The data transmitter transmits the data to the embedded module or the cloud platform through the laser communication line.

Preferably, the laser radar technology-based real-time monitoring and accurate temperature measurement are performed on the charging pile, and the method comprises the following steps:

(1) A laser radar transmitter transmits a laser signal to the charging pile;

(2) The laser radar receiver receives the laser signals reflected from the charging pile and converts the signals into digital signals;

(3) The image processor receives the digital signals, calculates the distance, speed and temperature information of the target object according to the reflection intensity and the phase difference of the signals, and generates corresponding images;

(4) And the display displays the image generated by the image processor and carries out early warning prompt according to the set threshold value.

Preferably, the intelligent identification and remote monitoring of the charging pile based on the RFID technology comprises the following steps:

(1) The RFID reader-writer periodically transmits a query signal to the RFID tag to activate the RFID tag;

(2) After receiving the query signal, the RFID tag dynamically measures the temperature of the charging pile based on the temperature sensor in the RFID tag, and sends the temperature data and the information of the charging pile to the RFID reader-writer as response signals;

(3) After receiving the response signal, the RFID reader-writer sends the data in the response signal to the embedded module or the cloud platform through the communication line of the Internet of things.

Preferably, the block chain technology and the intelligent contract technology based secure storage, verification and transmission of the data of the charging pile comprise the following steps:

(1) After receiving data sent by an RFID reader-writer, an acoustic wave sensor, a laser radar or an infrared thermal imager, the embedded module or the cloud platform analyzes and stores the data, and sends the data to other nodes through a blockchain network;

(2) After the block chain network receives the data, verifying the validity and legality of the data, and recording the data on the block chain;

(3) Other nodes acquire data through a blockchain network and synchronize and back up the data;

(4) And triggering corresponding intelligent contracts by the embedded module or the cloud platform according to the data content and the conditions, and automatically processing and maintaining the data.

Preferably, the deep learning algorithm is used for extracting, classifying and identifying the characteristics of the data of the charging pile, and the method comprises the following steps:

(1) After receiving data sent by an RFID reader-writer, an acoustic wave sensor, a laser radar or an infrared thermal imager, the embedded module or the cloud platform analyzes and stores the data, and sends the data as input into a deep learning model;

(2) After the deep learning model receives data, carrying out feature extraction, classification and identification, and carrying out matching and judgment according to a pre-trained fault mode to output fault type and degree;

(3) And after receiving the result output by the deep learning model, the embedded module or the cloud platform carries out fire risk assessment and outputs early warning level and emergency measures.

Preferably, the early warning prompt includes intelligent early warning of abnormal temperature rise, overtemperature or fire according to temperature data and fault modes, and early warning state and emergency measures are displayed through a user terminal or a monitoring center, and the method comprises the following steps:

(1) After receiving temperature data sent by an acoustic wave sensor, a laser radar or an infrared thermal imager, the embedded module or the cloud platform compares the temperature data with a set threshold value, and if the temperature exceeds the threshold value, the embedded module or the cloud platform judges that the temperature rises abnormally or exceeds the temperature;

(2) After receiving the fault type and degree output by the deep learning model, the embedded module or the cloud platform compares the fault type and degree with a set threshold value, and if the fault type or degree is found to exceed the threshold value, the embedded module or the cloud platform judges that the fault type or degree is fire;

(3) The embedded module or the cloud platform carries out fire risk assessment according to abnormal temperature rise, overtemperature or fire judgment results, and outputs early warning level and emergency measures;

(4) The user terminal or the monitoring center receives and displays the early warning state and the emergency measures, and executes corresponding power-off, water spraying fire extinguishing and fire extinguishing notification operations according to the selection of the user

The invention relates to a device applied to a charging early warning method of a charging pile, which comprises the following components:

The infrared thermal imager is used for carrying out high-definition monitoring and accurate temperature measurement on the charging pile;

the acoustic wave sensor is used for monitoring acoustic wave signals generated by the charging pile in real time and accurately measuring the temperature;

The laser radar monitors the charging pile in real time and measures the temperature accurately;

The RFID reader-writer is used for carrying out intelligent identification and remote monitoring on the charging pile;

The data transmitter is used for transmitting the acquired data to the embedded module or the cloud platform through a laser communication line or an Internet of things communication line;

the embedded module or cloud platform performs data storage, verification and transmission based on a blockchain technology and an intelligent contract technology, performs data analysis and fault identification based on a deep learning algorithm, and outputs early warning states and emergency measures;

and the user terminal or the monitoring center receives and displays the early warning state and the emergency measures, and executes corresponding operations according to the selection of the user.

Compared with the prior art, the invention has the following beneficial effects:

1. According to the invention, the infrared thermal imaging technology, the acoustic wave sensor technology and the laser radar technology are utilized to carry out high-definition monitoring and accurate temperature measurement on the charging pile, so that the inside and the outside of the charging pile can be covered, and the omnibearing monitoring is realized;

2. The invention uses RFID technology to carry out intelligent identification and remote monitoring on the charging pile, and can realize wireless, passive, low-cost, low-energy consumption, long service life, high reliability, high safety and high efficiency monitoring;

3. The invention utilizes the blockchain technology and the intelligent contract technology to safely store, verify and transmit the data of the charging pile, and can realize the data processing with decentralization, non-falsification, traceability, credibility, automatic execution and low cost;

4. The invention utilizes the deep learning algorithm to extract, classify and identify the characteristics of the data of the charging pile, and can realize fault identification with high precision, high efficiency, high intelligence, high adaptability and high sensitivity;

5. According to the invention, the abnormal temperature rise, overtemperature or fire disaster intelligent early warning is carried out according to the temperature data and the fault mode, and the early warning state and the emergency measures are displayed through the user terminal or the monitoring center, so that timely, accurate, detailed, friendly and convenient early warning prompt can be realized.

Drawings

FIG. 1 is a schematic illustration of the structure of the present invention;

FIG. 2 is a schematic diagram of an RFID tag of the present invention;

FIG. 3 is a schematic diagram of a blockchain system of the present invention;

FIG. 4 is a schematic diagram of a smart contract system of the present invention;

FIG. 5 is a schematic diagram of a deep learning model of the present invention;

FIG. 6 is a general flow chart of a charging early warning method of the charging pile of the present invention;

FIG. 7 is a flow chart of monitoring and early warning of a charging pile according to the infrared thermal imaging technique of the invention;

FIG. 8 is a flow chart of monitoring and early warning of a charging pile according to the acoustic wave sensor technology of the present invention;

FIG. 9 is a flow chart of monitoring and early warning of a charging pile of the laser radar technology of the invention;

FIG. 10 is a flow chart of identification and monitoring of a charging pile of the RFID technology of the present invention;

FIG. 11 is a flow chart of a charge stake data processing for the blockchain technique and the smart contract technique of the present invention;

FIG. 12 is a flow chart of fault recognition and early warning of the charging pile according to the deep learning algorithm of the invention;

FIG. 13 is a flow chart of fire risk assessment and emergency measures for the charging stake of the present invention;

Fig. 14 is a schematic diagram of a device structure of the charging early warning method of the present invention.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

Referring to fig. 1 to 14, the present invention provides a technical solution: the charging early warning method of the charging pile comprises the following steps:

s101: carry out real-time supervision to charging pile, including carrying out high definition control and accurate temperature measurement to charging pile based on infrared thermal imaging technique, acoustic wave sensor technique and laser radar technique to and carry out intelligent identification and remote monitoring to charging pile based on RFID technique.

In the step, the invention utilizes the infrared thermal imaging technology, the acoustic wave sensor technology and the laser radar technology to carry out high-definition monitoring and accurate temperature measurement on the charging pile, and can cover the inside and the outside of the charging pile to realize omnibearing monitoring. Specifically:

High-definition monitoring and accurate temperature measurement are carried out on charging piles based on infrared thermal imaging technology: according to the invention, the plurality of infrared thermal imagers are arranged around the charging pile and are used for carrying out high-definition monitoring and accurate temperature measurement on the charging pile. The infrared thermal imager emits infrared rays to the charging pile and receives the infrared rays reflected from the charging pile. The infrared thermal imager converts the received infrared rays into digital signals, calculates information such as the temperature of a target object according to the intensity and the wavelength of the signals, and generates corresponding images. And displaying an image generated by the infrared thermal imager by the display, and carrying out early warning prompt according to the set threshold value. The infrared thermal imaging technology can monitor the surface temperature of the charging pile in real time, without contact, damage, high resolution, high sensitivity and high reliability, and can effectively find and locate the abnormal temperature rise or overtemperature region of the charging pile.

Real-time monitoring and accurate temperature measurement are carried out on the charging pile based on the acoustic wave sensor technology: the invention is provided with a plurality of acoustic wave sensors in the charging pile, and the acoustic wave sensors are used for monitoring acoustic wave signals generated by the charging pile in real time and accurately measuring the temperature. The sound wave sensor transmitter transmits sound wave signals to the interior of the charging pile. The acoustic wave sensor receiver receives acoustic wave signals reflected from the inside of the charging pile and converts the signals into digital signals. The data processor receives the digital signals, calculates information such as the temperature of the target object according to the propagation time and frequency of the signals, and generates corresponding data. The data transmitter transmits the data to the embedded module or the cloud platform through the laser communication line. The acoustic wave sensor technology can monitor the internal temperature of the charging pile in real time, without contact, damage, high resolution, high sensitivity and high reliability, and can effectively find and locate the abnormal temperature rise or overtemperature region of the charging pile.

According to the invention, a plurality of laser radars are arranged around the charging pile and are used for monitoring the charging pile in real time and accurately measuring the temperature; lidar technology uses a laser to measure the distance and reflectivity of an object and converts it into a point cloud that displays shape and surface characteristics.

The laser radar transmitter transmits laser signals to the charging pile, and the laser radar receiver receives the laser signals reflected from the charging pile and converts the signals into digital signals. The image processor receives the digital signals, calculates information such as the distance, the speed, the temperature and the like of the target object according to the reflection intensity and the phase difference of the signals, and generates corresponding images. And the display displays the image generated by the image processor and carries out early warning prompt according to the set threshold value. The laser radar technology can monitor the surface temperature of the charging pile in real time, in a non-contact, non-destructive, high-resolution, high-sensitivity and high-reliability manner, and effectively discover and locate the abnormal temperature rise or overtemperature region of the charging pile.

In the step, the invention also utilizes the RFID technology to carry out intelligent identification and remote monitoring on the charging pile, and can realize wireless, passive, low-cost, low-energy consumption, long service life, high reliability, high safety and high efficiency monitoring.

One possible way to monitor the temperature of the charging pile using these techniques is as follows:

An infrared camera is mounted on each charging post or a central station covering a plurality of charging posts. The infrared camera may capture thermal images of the charging stake at fixed time intervals or triggered by certain events (e.g., user request or abnormal conditions). Thermal imaging may show hot or cold spots on the charging stake indicating a potential problem or failure.

An acoustic wave sensor is installed on each charging post or a central station covering a plurality of charging posts. The acoustic wave sensor may capture the acoustic wave of the charging stake at fixed time intervals or triggered by some event, such as a user request or an abnormal condition. The sound waves may show a change in frequency or amplitude on the charging stake, indicating a change in temperature or fluctuation.

A lidar sensor is mounted on each charging post or a central station covering a plurality of charging posts. Lidar sensors may capture the point cloud of the charging stake at fixed time intervals or triggered by certain events (e.g., user request or abnormal conditions). The point cloud may display the shape and surface characteristics of the charging stake, indicating thermal deformation or damage.

Thus, a comprehensive, reliable and accurate temperature measurement system of the charging pile can be realized;

the temperature fluctuation score and the temperature convexity score are two indexes which we propose for evaluating the temperature condition of the charging pile. The temperature fluctuation score reflects the degree of change of the temperature in the charging pile area, and the larger the temperature fluctuation score is, the more unstable the temperature is; the temperature convexity score reflects the distribution shape of the temperature in the charging pile area, and the larger the temperature is, the more the convexity distribution is presented. We use some innovative calculation methods and formulas to calculate these two indices in order to more accurately detect and prevent the temperature anomalies and fire risks of the charging stake.

wherein t _i,j represents the temperature of the jth sub-zone in the ith round of detection, The average temperature of all subareas in the ith round of detection is represented, and n1 represents the number of subareas. The temperature fluctuation score reflects the degree of change of the temperature in the charging pile area, and the larger the temperature fluctuation score is, the more unstable the temperature is;

The temperature convexity score cfs _i is calculated by the following formula:

Wherein t _i,j represents the temperature of the jth sub-region in the ith round of detection, t _i represents the average temperature of all the sub-regions in the ith round of detection, and n ₁ represents the number of sub-regions. The temperature convexity score reflects the degree of peaks or valleys of the temperature within the charging pile area, with a greater value indicating a less smooth temperature.

Thus, the temperature detection score of each subarea can be determined according to the temperature fluctuation score and the temperature convexity score, and the subarea with abnormal temperature can be determined according to the temperature detection score.

To calculate the temperature early warning score, the following formula is used:

Wherein tws _i represents a temperature early warning score in the ith round of detection, t _i,j represents the temperature of the jth sub-region in the ith round of detection, t _i represents the average temperature of all sub-regions in the ith round of detection, t _th represents a temperature preset threshold value, and n1 represents the number of sub-regions. The temperature early warning score reflects the degree that the temperature in the charging pile area exceeds a threshold value, and the larger the temperature is, the higher the temperature is.

To calculate the temperature overrun score, we can use the following formula:

Wherein tos _i represents the temperature overrun score in the ith round of detection, Representing partial derivatives of the temperature of the jth sub-region in the x-direction and the y-direction in the ith round of detection, and n ₂ represents the number of sub-regions with the temperature greater than a preset threshold. The temperature overrun score reflects the rate of temperature change within the charging pile area, with a larger value indicating a more non-uniform temperature.

And determining whether a high-temperature risk exists in the charging pile area according to the temperature early warning score and the temperature overrun score.

To integrate these technologies into a smart charge pile monitoring and early warning system, an embedded module or cloud platform may be used to collect, process and analyze the data generated by these technologies. The embedded module or cloud platform can be connected with the infrared camera, the sound wave sensor and the laser radar sensor on each charging pile or central station through wireless or wired communication lines, and receive data sent by the infrared camera, the sound wave sensor and the laser radar sensor periodically or on demand. The embedded module or cloud platform can also interact with the blockchain network and the deep learning model, store, verify and transmit data to the blockchain, and input the data to the deep learning model for feature extraction, classification and identification. The embedded module or the cloud platform can also communicate with the user terminal or the monitoring center, send and display the data analysis result and the early warning information to the user, and execute corresponding operation according to the selection of the user.

Therefore, an integrated, intelligent and automatic charging pile monitoring and early warning system can be realized, the temperature abnormality and fire risk of the charging pile can be effectively detected and prevented, and the safety and reliability of the charging pile are improved.

Specifically, the invention installs an RFID tag on each charging pile for intelligent identification and remote monitoring of the charging pile. The RFID tag includes an antenna for receiving and transmitting radio waves and a chip for storing and processing data. The RFID tag is also internally provided with a temperature sensor for dynamically measuring the temperature of the charging pile. The RFID reader-writer periodically transmits a query signal to the RFID tag to activate the RFID tag. After receiving the inquiry signal, the RFID tag dynamically measures the temperature of the charging pile based on the temperature sensor in the RFID tag, and sends the temperature data and the information of the charging pile to the RFID reader-writer together as a response signal. After receiving the response signal, the RFID reader-writer sends the data in the response signal to the embedded module or the cloud platform through the communication line of the Internet of things. The RFID technology can monitor basic information and temperature data of the charging pile in real time, wirelessly, passively, at low cost, with low energy consumption, long service life, high reliability, high safety and high efficiency, and effectively realize intelligent identification and remote monitoring of the charging pile.

S102: and carrying out data analysis on the data of the charging pile, including safe storage, verification and transmission of the data of the charging pile based on a blockchain technology and an intelligent contract technology, and feature extraction, classification and identification of the data of the charging pile based on a deep learning algorithm.

In the step, the invention utilizes the blockchain technology and the intelligent contract technology to safely store, verify and transmit the data of the charging pile, and can realize the data processing with decentralization, non-falsification, traceability, credibility, automatic execution and low cost.

Specifically: the invention establishes a distributed database system based on a blockchain technology, which is used for safely storing, verifying and transmitting the data of the charging pile. The system is composed of a plurality of nodes, each node is a separate server or computer, and the nodes are connected through a network and follow the same common protocol. Each node stores a complete set of blockchain data, which is a chain data structure formed by linking a plurality of blocks in time sequence. Each block contains the data of a group of charging piles, and the hash value of the previous block and the hash value of the current block, thereby realizing chain connection between the blocks. The data of each block is encrypted and signed, so that the safety and the integrity of the data are ensured. The generation of each block needs to be subjected to certain calculation difficulty, so that the stability and the non-tamper property of the block chain are ensured. Each node can participate in the generation and verification of the block, and a decentralization and consensus mechanism of the block chain is realized. The invention also utilizes intelligent contract technology to automatically process and maintain the data of the charging pile. The intelligent contract is a programmable contract based on the blockchain technology, and can automatically execute corresponding operations according to preset conditions and rules. The invention deploys a plurality of intelligent contracts on the blockchain for automatically processing and maintaining the data of the charging pile.

Among other things, to ensure the security and integrity of the data collected by the charging stake, blockchain technology and smart contract technology are used to store, verify and transmit the data. Blockchains are a distributed ledger that records transactions in a point-to-point network, using cryptography, consensus mechanisms, and smart contracts to provide a new trust system. A smart contract is a self-executing program that is stored on a blockchain and runs when predefined conditions are met. They can automatically implement agreements and behaviors between participants without the need for intermediaries or central authorities.

One possible way of charging pile data management using blockchain and smart contract technology is as follows:

each charging stake has a unique identity and a corresponding smart contract defining its attributes such as location, price, availability, etc.

Each charging stake is connected to a blockchain network that can communicate with other nodes (e.g., other charging stakes, users, service providers, etc.) and record its data (e.g., power consumption, temperature, status, etc.) on the blockchain.

The data recorded on the blockchain is encrypted, non-tamperable, and can be verified by all nodes. They may also be used to trigger functions of smart contracts such as billing, maintenance, rewards, and the like.

Users may access the blockchain network through their mobile devices or vehicles, they may search for available charging posts, reservations, payment services, evaluation quality, etc. through smart contracts.

The service provider may also access the blockchain network, monitor the performance of the charging stake, provide maintenance and support, collect fees, etc., also through smart contracts.

Thus, a safe, transparent and efficient data management system for the charging pile can be realized.

The intelligent contract technology is a programmable contract based on the blockchain technology, and can automatically execute corresponding operations according to preset conditions and rules. The intelligent contract technology mainly comprises the following aspects:

intelligent contract: the intelligent contract is a programmable contract based on the blockchain technology, and can automatically execute corresponding operations according to preset conditions and rules. Intelligent contracts consist of code for defining the logic and functions of the contract and data for storing the state and variables of the contract. The smart contracts may be deployed onto the blockchain as part of the blockchain, thereby enabling the non-tamperable, traceable, trusted, automated execution and low cost features of the smart contracts.

Intelligent contract programming language: the intelligent contract programming language is a special programming language for writing intelligent contracts, and has the characteristics of high efficiency, safety, easiness in use and the like. Intelligent contract programming languages are typically required to support the following aspects:

Data type and structure: smart contract programming languages need to support basic data types and structures, such as integers, floating point numbers, boolean values, strings, arrays, mappings, etc., in order to store and process data.

Control flow and function: intelligent contract programming languages need to support basic control flows and functions such as conditional decisions, loops, recursions, return values, etc. in order to implement logic and functionality.

Event and message: smart contract programming languages need to support events and messages in order to enable interactions and communications with other smart contracts or external systems.

Security and exception handling: the smart contract programming language needs to support security and exception handling in order to prevent malicious attacks or unexpected errors and to provide a corresponding recovery mechanism.

Optimizing and testing: smart contract programming languages need to support optimization and testing in order to improve the performance and quality of smart contracts and reduce execution costs and risks. Currently, common smart contract programming languages include Solidity, vyper, serpent. An intelligent contract platform: the intelligent contract platform is an operating environment based on the blockchain technology and is used for deploying and executing intelligent contracts. An intelligent contract platform is generally required to provide several aspects:

Virtual machine: a virtual machine is a software system that emulates computer hardware for executing the bytecodes or instructions of a smart contract. The virtual machine needs to ensure correctness, security, isolation and compatibility of execution.

Library and interface: libraries and interfaces are a set of codes or protocols that provide common functions or services for assisting or expanding the functionality of smart contracts. Libraries and interfaces may include mathematical operations, encryption algorithms, random number generation, time stamp retrieval, data storage, network communications, and the like.

Tool and frame: tools and frameworks are a set of software or systems that provide auxiliary functions or services for developing or managing smart contracts. Tools and frameworks can include compilers, debuggers, testers, deployers, monitors, analyzers, and the like. Currently, common intelligent contract platforms include ethernet, EOS, HYPERLEDGER FABRIC, and the like. In the present invention, the following technical details and principles are utilized: blockchain techniques: the invention adopts Ethereum as a platform of a blockchain technology, ethereum is a public chain based on a PoW consensus protocol, supports a complete Solidity of Turing as an intelligent contract programming language, and provides Ethereum Virtual Machines (EVM) as an execution environment. Ethereum also provides web3.js as an interface to interact with external systems, and Truffle, ganache et al as a tool and framework for development and testing. Intelligent contract technology: according to the invention, a plurality of intelligent contracts are deployed on Ethereum and are used for automatically processing and maintaining the data of the charging pile.

Specifically, the present invention deploys several smart contracts:

A data storage contract: the data storage contract is used for storing basic information and temperature data of the charging pile, and early warning states and emergency measures. The data storage contract adopts mapping (mapping) as a data structure, the number of the charging pile is used as a key, and the information and data of the charging pile are used as values, so that efficient data access and updating are realized.

Data validation contracts: the data verification contract is used for verifying the validity and legality of the data of the charging pile so as to prevent a malicious node or an attacker from forging or falsifying the data. The data verification contract adopts a hash function and a digital signature as verification methods, carries out hash operation on the data of the charging pile, and signs the data by using the private key of the charging pile, thereby realizing the integrity and identity authentication of the data.

A data transfer contract: the data transmission contract is used for transmitting the data of the charging pile to other nodes or systems so as to realize sharing and utilization of the data. The data transmission contract adopts an event (event) as a transmission method, takes the data of the charging pile as the parameter of the event, and uses an init statement to trigger the event, thereby realizing the broadcasting and notification of the data.

A data processing contract: the data processing contract is used for carrying out fire risk assessment according to the data of the charging pile and outputting early warning grade and emergency measures. The data processing contract adopts condition judgment and function call as processing methods, compares the temperature data and fault mode of the charging pile with a set threshold value, and calls a corresponding function according to a comparison result, thereby realizing intelligent early warning of abnormal temperature rise, overtemperature or fire.

Data maintenance contracts: the data maintenance contract is used to perform corresponding operations such as cutting off power, spraying fire, and notifying fire according to the user's selection. The data maintenance contract adopts a message (message) as an execution method, takes the selection of a user as a parameter of the message, and sends the message to corresponding equipment or a system by using a send statement, thereby realizing the automatic execution of emergency measures. The block chain technology and the intelligent contract technology can safely store, verify and transmit the data of the charging pile, and automatically process and maintain the data, so that the safety, reliability, efficiency and intelligence of the charging pile monitoring and early warning system are effectively improved. In the step, the invention also utilizes the deep learning algorithm to extract, classify and identify the characteristics of the data of the charging pile, and can realize fault identification with high precision, high efficiency, high intelligence, high adaptability and high sensitivity.

Specifically: in the process of extracting, classifying and identifying the characteristics of the data of the charging pile based on the deep learning algorithm, the invention deploys a deep learning model on the embedded module or the cloud platform, and the deep learning model is used for extracting, classifying and identifying the characteristics of the data of the charging pile. The deep learning model is a machine learning model for realizing complex functions or tasks by using a multi-layer neural network, and has strong expression capability and learning capability. After the deep learning model receives data sent by the RFID reader-writer, the acoustic wave sensor, the laser radar or the infrared thermal imager, the deep learning model performs feature extraction, classification and identification after receiving the data sent by the RFID reader-writer, the acoustic wave sensor, the laser radar or the infrared thermal imager, performs matching and judgment according to a pre-trained fault mode, and outputs fault type and degree. The deep learning algorithm can perform fault identification with high accuracy, high efficiency, high intelligence, high adaptability and high sensitivity on the data of the charging pile, and can effectively find and locate abnormal temperature rise, overtemperature or fire disaster areas of the charging pile.

In order to extract, classify, and identify features of the charging pile data, deep learning algorithms are used to learn and generate high-level representations from the data. Deep learning does not require manual design of features, but rather automatically filters and extracts high-dimensional features from the data.

One way to perform the charging pile data analysis using the deep learning algorithm is as follows:

Convolutional Neural Networks (CNNs) may be used to process image data of the charging posts, such as infrared thermal imaging, acoustic imaging, or lidar imaging. CNNs consist of multiple layers of filters that convolve with the input image and produce feature maps, capturing local patterns and structures. CNNs can learn to extract features, such as edges, shapes, textures, colors, etc., from low-level to high-level.

The sequential data of the charging posts, such as the power consumption time series or the temperature change time series, may be processed using a Recurrent Neural Network (RNN) or a long short term memory network (LSTM). RNN and LSTM are composed of multiple units with internal memory that can process sequence inputs one by one. They may learn to extract features such as trends, periods, anomalies, etc. from the time series data.

The fully connected or softmax layer is used to classify and identify CNN or RNN/LSTM extracted features. The fully connected layer is made up of a plurality of neurons that connect each input to each output. The Softmax layer is a special case of a fully connected layer that outputs probability distributions for a set of classes. They may learn to map features to different categories or labels, such as normal/abnormal status, fault type, temperature range, etc.

Thus, a robust, accurate and self-adaptive charging pile data analysis system can be realized.

The invention adopts the convolutional neural network (Convolutional Neural Network, CNN) as a deep learning model, and the CNN is a neural network for realizing feature extraction and classification by using convolutional operation, and has the characteristics of high efficiency, robustness, expandability and the like.

CNN mainly includes the following aspects:

convolution layer: the convolution layer is a core layer of the CNN and is used for carrying out convolution operation on input data and extracting local characteristics of the data. The convolution layer is made up of a plurality of convolution kernels, each having a filter and a bias term. The convolution layer generates a corresponding feature map by sliding a convolution kernel over the input data and performing element-by-element multiply-add operations. The number of the convolution layers can be multiple, and each convolution layer can be provided with different parameters such as the number, the size, the step length, the filling and the like of the convolution kernels so as to realize different characteristic extraction effects.

Pooling layer: the pooling layer is a dimension reduction layer of the CNN and is used for performing pooling operation on output data of the convolution layer and reducing the dimension and complexity of the data. The pooling layer divides the input data into a plurality of sub-areas, and performs maximum value, average value or other statistical operation on each sub-area to generate a corresponding pooling result. The pooling layers can be multiple, and each pooling layer can be provided with parameters such as different pooling types, sizes, step sizes, filling and the like so as to realize different dimension reduction effects.

Full tie layer: the full connection layer is an output layer of the CNN and is used for carrying out final processing on output data of the pooling layer and outputting a result. The fully connected layer is composed of a plurality of neurons, each neuron is connected with all neurons of the previous layer, and has a weight vector and a bias term. The full connection layer performs dot multiplication operation on the input data and the weight vector, and adds a bias term to generate a corresponding output result. The number of the full-connection layers can be multiple, and each full-connection layer can be provided with different parameters such as the number of neurons, an activation function and the like so as to achieve different processing effects.

Activation function: the activation function is a function for increasing the nonlinearity and expressive power of the CNN, and is used to perform nonlinear transformation on the output data of the convolutional layer or the fully-connected layer. The activation function may be varied, such as Linear (Linear), sigmoid, tanh, reLU, leak ReLU, ELU, etc.

Loss function: the loss function is a function for measuring the difference or error between the CNN output result and the real result, and is used for guiding the optimization and updating of the CNN. The loss function may be varied, such as Mean Square Error (MSE), cross Entropy (CE), contrast Loss (CL), and the like.

Optimization algorithm: the optimization algorithm is an algorithm for adjusting CNN parameters to minimize loss function values, for improving the performance and quality of CNN. There are various optimization algorithms, such as Gradient Descent (GD), random gradient descent (SGD), momentum Method (MM), adaptive gradient (AdaGrad), adaptive momentum estimation (Adam), etc.

The invention adopts a CNN model composed of four convolution layers, two pooling layers and two full-connection layers, and is used for extracting, classifying and identifying the characteristics of the data of the charging pile. Specifically, the present invention employs the following parameter settings:

First convolution layer: the first convolution layer uses 32 convolution kernels with the size of 3×3, the step size is 1, the padding is same, and the activation function is ReLU;

second convolution layer: the second convolution layer uses 64 convolution kernels of 3×3 size, the step size is 1, the padding is same, and the activation function is ReLU;

A first pooling layer: the first pooling layer uses maximum pooling, the pooling size is 2 multiplied by 2, and the step length is 2;

the third convolution layer uses 128 convolution kernels of 3×3 size, the step size is 1, the padding is same, and the activation function is ReLU;

Fourth convolution layer: the fourth convolutional layer uses 256 convolutional kernels of 3×3 size, the step size is 1, the padding is same, and the activation function is ReLU;

A second pooling layer: the second pooling layer uses maximum pooling, the pooling size is 2×2, and the step length is 2;

First full tie layer: the first fully-connected layer uses 1024 neurons, and the activation function is ReLU;

Second full tie layer: the second fully-connected layer uses 4 neurons, and the activation function is Softmax and is used for outputting probability distribution of fault type and degree;

Loss function: the loss function uses Cross Entropy (CE) as a function for measuring the difference or error between the CNN output result and the real result, and is used for guiding the optimization and updating of CNN;

Optimization algorithm: the optimization algorithm uses adaptive momentum estimation (Adam) as an algorithm to adjust CNN parameters to minimize loss function values for improving CNN performance and quality. The CNN model can perform fault identification with high accuracy, high efficiency, high intelligence, high adaptability and high sensitivity on the data of the charging pile, and can effectively find and locate abnormal temperature rise, overtemperature or fire disaster areas of the charging pile.

S103: and carrying out early warning prompt according to the temperature data and the fault mode, namely carrying out intelligent early warning of abnormal temperature rise, overtemperature or fire disaster according to the temperature data and the fault mode, and displaying early warning states and emergency measures through a user terminal or a monitoring center. In the step, the invention carries out intelligent early warning of abnormal temperature rise, overtemperature or fire disaster according to the temperature data and the fault mode, and displays early warning states and emergency measures through the user terminal or the monitoring center, thereby realizing timely, accurate, detailed, friendly and convenient early warning prompt. Specifically: according to the temperature data and the fault mode, carrying out intelligent early warning on abnormal temperature rise, overtemperature or fire disaster: according to the invention, a data processing contract is deployed on the embedded module or the cloud platform, and the data processing contract is used for carrying out fire risk assessment according to the temperature data and the fault mode of the charging pile and outputting early warning grades and emergency measures. The data processing contract adopts condition judgment and function call as processing methods, compares the temperature data and fault mode of the charging pile with a set threshold value, and calls a corresponding function according to a comparison result, thereby realizing intelligent early warning of abnormal temperature rise, overtemperature or fire. Specifically, the data processing contract employs the following logic and rules:

if the temperature data of the charging pile is lower than the set normal temperature threshold value, judging that the charging pile is in a normal state, and carrying out early warning is not needed;

if the temperature data of the charging pile is higher than the set normal temperature threshold value but lower than the set abnormal temperature rise threshold value, judging the charging pile to be in an abnormal temperature rise state, and calling an abnormal temperature rise early warning function;

if the temperature data of the charging pile is higher than the set abnormal temperature rise threshold value but lower than the set overtemperature threshold value, judging that the charging pile is in an overtemperature state, and calling an overtemperature early warning function;

If the temperature data of the charging pile is higher than the set overtemperature threshold, judging a fire state, and calling a fire early warning function;

If the fault mode of the charging pile is matched with the set fault mode threshold, judging the fire state, and calling a fire early warning function. Abnormal temperature rise early warning function: the abnormal temperature rise early warning function is used for outputting early warning grades and emergency measures of abnormal temperature rise. The abnormal temperature rise early warning function outputs early warning grade as yellow according to the temperature data and the fault mode of the charging pile, and outputs emergency measures for checking a connecting line and a power switch of the charging pile and reducing charging power and time. Overtemperature early warning function: the overtemperature early warning function is used for outputting early warning grades and emergency measures of overtemperature. The overtemperature early warning function outputs an early warning grade to be orange according to the temperature data and the fault mode of the charging pile, and outputs emergency measures to cut off the power supply of the charging pile, and ventilate and cool down and check a heat dissipation system of the charging pile. Fire early warning function: the fire early warning function is used for outputting early warning level and emergency measures of the fire. The fire disaster early warning function outputs early warning grade red according to the temperature data and the fault mode of the charging pile, and outputs emergency measures for cutting off the power supply of the charging pile, spraying water to extinguish fire and notifying fire. Displaying early warning states and emergency measures through a user terminal or a monitoring center: the invention deploys a data display contract on the user terminal or the monitoring center, which is used for receiving and displaying the early warning state and the emergency measure output by the data processing contract, and executing corresponding operation according to the selection of the user. The data display contract adopts a message (message) as a display method, the early warning state and the emergency measure output by the data processing contract are used as parameters of the message, and the message is sent to the user terminal or the monitoring center by using a send statement, so that the display of the early warning state and the emergency measure is realized. The data display contract also adopts a message (message) as an execution method, takes the selection of a user as a parameter of the message, and sends the message to corresponding equipment or a system by using a send statement, so that the execution of emergency measures is realized. Specifically, the data display contract employs the following logic and rules:

if the normal state of the data processing contract output is received, not performing any operation;

If an abnormal temperature rise state output by the data processing contract is received, displaying a yellow early warning signal on the user terminal or the monitoring center, prompting a user to check a connecting line and a power switch of the charging pile, and reducing charging power and time;

If the overtemperature state output by the data processing contract is received, displaying an orange early warning signal on a user terminal or a monitoring center, prompting a user to cut off a power supply of the charging pile, and ventilating and cooling and checking a heat dissipation system of the charging pile;

If the fire state output by the data processing contract is received, displaying a red early warning signal on a user terminal or a monitoring center, prompting a user to cut off the power supply of the charging pile, spraying water to extinguish the fire and notifying the fire;

if the user chooses to perform the corresponding operation, the user's choice is sent as a message to the corresponding device or system, such as cutting off the power to the charging pile, spraying water to extinguish the fire, and notifying the fire. The early warning state and emergency measures are displayed through the user terminal or the monitoring center, so that early warning prompt which is timely, accurate, detailed, friendly and convenient can be realized, and the safety awareness and coping capacity of the user are effectively enhanced.

Example 1: intelligent charging station for electric bus

In this embodiment, the present invention is applied to an intelligent charging station for an electric bus, which can monitor and early warn of the charge states of a plurality of charging posts at the same time and optimize the charging efficiency and safety of the electric bus.

The intelligent charging station consists of a plurality of charging piles, wherein each charging pile is provided with an RFID tag, an infrared thermal imager, an acoustic wave sensor and a laser radar. The RFID tag is used to remotely identify and monitor the charging stake. Thermal infrared imagers, acoustic wave sensors, and lidar are used to monitor and measure the temperature inside and outside the charging stake. The intelligent charging station is also provided with an embedded module or a cloud platform, and is connected with the charging pile through a wireless or wired communication line. The embedded module or cloud platform is provided with a blockchain system, an intelligent contract system and a deep learning model. The blockchain system is used to securely store, verify and transmit the data of the charging stake. The smart contract system is used to automatically process and maintain data for the charging stake. The deep learning model is used to extract, classify, and identify features of the data of the charging stake.

The workflow of this intelligent charging station is as follows:

When an electric bus arrives at the intelligent charging station, it scans for and establishes a connection with an RFID tag of an empty charging post. And the RFID tag sends the basic information and temperature data of the charging pile to the embedded module or the cloud platform through the RFID reader-writer.

The embedded module or the cloud platform receives the data of the charging pile and verifies the validity and legality of the charging pile through a data verification contract. It then sends the charging stake data to other nodes or systems via a data transfer contract.

The embedded module or cloud platform also sends a query signal to the infrared thermal imager, acoustic sensor and lidar of the charging pile and activates them. The infrared thermal imager, the acoustic wave sensor and the laser radar monitor and measure the temperature inside and outside the charging pile, and send the data to the embedded module or the cloud platform through the laser communication line.

The embedded module or the cloud platform receives the data of the infrared thermal imager, the acoustic sensor and the laser radar, and verifies the validity and the legality of the infrared thermal imager, the acoustic sensor and the laser radar through a data verification contract. It then sends their data to other nodes or systems via a data transfer contract.

The embedded module or the cloud platform also inputs the data of all the charging piles into the deep learning model, performs feature extraction, classification and identification, performs matching and judgment according to a pre-trained fault mode, and outputs fault type and degree.

And the embedded module or the cloud platform carries out fire risk assessment through a data processing contract according to the fault type and degree output by the deep learning model, and outputs an early warning level and emergency measures.

The embedded module or the cloud platform sends and displays the early warning level and the emergency measure to the user terminal or the monitoring center through the data display contract, and corresponding operation is executed according to the selection of the user.

The user terminal or the monitoring center receives and displays the early warning level and the emergency measures and prompts the user to execute corresponding operations, such as cutting off the power supply of the charging pile, spraying water to extinguish fire and notifying fire.

The user terminal or the monitoring center sends the user selection to the embedded module or the cloud platform, and the embedded module or the cloud platform sends the user selection to corresponding equipment or systems through a data maintenance contract, such as cutting off the power supply of the charging pile, spraying water to extinguish fire and notifying fire.

The intelligent charging station can monitor and early warn the charging states of a plurality of charging piles simultaneously, optimize the charging efficiency and the safety of the electric bus, and effectively improve the safety, the reliability, the efficiency and the intelligence of the charging pile monitoring and early warning system.

In order to verify the effect of the invention, a comparison experiment is performed in the embodiment, and the invention is compared with a traditional charging pile monitoring and early warning system. The traditional charging pile monitoring and early warning system mainly uses a temperature sensor to monitor the temperature of a charging pile and uses simple threshold judgment to carry out early warning prompt.

The results of the comparative experiments are shown in the following table:

from the table, the invention is superior to the traditional system in terms of fault recognition accuracy, efficiency, early warning prompt timeliness, detail, friendliness and convenience, and has obvious advantages and innovations.

Example 2: intelligent charging pile for electric automobile

In this embodiment, the present invention is applied to an intelligent charging pile for an electric vehicle, which can automatically adjust charging power and time according to charging requirements of the electric vehicle and power supply conditions of a power grid, and realize settlement and payment of charging fees.

The intelligent charging pile consists of a charging socket, an RFID tag, an infrared thermal imager, an acoustic wave sensor and a laser radar. The charging socket is used for being connected with a charging interface of the electric automobile and providing electric energy for the charging interface. The RFID tag is used to identify and monitor the charging stake. The infrared thermal imager, the acoustic wave sensor and the laser radar are used for monitoring and measuring the temperature of the charging pile. The intelligent charging pile is also connected with an embedded module or a cloud platform and is communicated with the embedded module or the cloud platform through a wireless or wired communication line. The embedded module or cloud platform is provided with a blockchain system, an intelligent contract system and a deep learning model. The blockchain system is used to securely store, verify and transmit the data of the charging stake. The intelligent contract system is used for automatically processing and maintaining data of the charging pile and realizing settlement and payment of charging fees. The deep learning model is used for extracting, classifying and identifying the characteristics of the data of the charging pile and predicting the charging requirement of the electric automobile and the power supply condition of the power grid.

The working flow of the intelligent charging pile is as follows:

When an electric automobile needs to be charged, the RFID card of the electric automobile is placed on the RFID reader-writer of the charging pile so as to verify the identity and balance of the electric automobile. If the verification is successful, the charging pile unlocks the charging socket and prompts the user to insert the charging interface.

When a user inserts the charging interface of the charging pile, the charging pile starts to provide low-power initial charging for the user, and basic information and temperature data of the charging pile are sent to the embedded module or the cloud platform through the RFID tag.

The embedded module or cloud platform also sends a query signal to the infrared thermal imager, the acoustic sensor, and the lidar and activates them. The infrared thermal imager, the acoustic wave sensor and the laser radar monitor and measure the temperature of the charging pile and send the data to the embedded module or the cloud platform through the laser communication line.

The embedded module or the cloud platform also inputs the data of all the charging piles into the deep learning model, performs feature extraction, classification and identification, performs matching and judgment according to a pre-trained fault mode, and outputs fault type and degree. Meanwhile, the deep learning model predicts the charging requirement of the electric automobile and the power supply condition of the power grid according to the charging history data and the current data of the electric automobile and the supply and demand data and the price data of the power grid, and outputs optimal charging power and time.

And the embedded module or the cloud platform carries out fire risk assessment through a data processing contract according to the fault type and degree output by the deep learning model, and outputs an early warning level and emergency measures. Meanwhile, the embedded module or the cloud platform adjusts the charging power and time of the charging pile through the data adjustment contract according to the optimal charging power and time output by the deep learning model, and outputs charging cost.

The embedded module or the cloud platform sends and displays the early warning grade, the emergency measure, the charging power, the charging time and the charging cost to the user terminal or the monitoring center through the data display contract, and executes corresponding operation according to the selection of the user.

The user terminal or the monitoring center receives and displays the early warning level, the emergency measure, the charging power, the charging time and the charging cost, and prompts the user to execute corresponding operations, such as cutting off the power supply of the charging pile, spraying water to extinguish fire, notifying fire, confirming the charging cost and the like.

The user terminal or the monitoring center sends the user selection to the embedded module or the cloud platform, and the embedded module or the cloud platform sends the user selection to corresponding equipment or a system through a data maintenance contract, such as cutting off the power supply of the charging pile, spraying water to extinguish fire, notifying fire and the like. Meanwhile, the embedded module or the cloud platform sends the selection of the user to a corresponding payment platform, such as WeChat, payment treasures and the like, through a data payment contract, and settlement and payment of charging fees are realized.

The intelligent charging pile can automatically adjust charging power and time according to charging requirements of the electric automobile and power supply conditions of a power grid, and can realize settlement and payment of charging cost, so that safety, reliability, efficiency and intelligence of a charging pile monitoring and early warning system are effectively improved.

In order to verify the effect of the present invention, a comparative experiment was performed in this example, and the present invention was compared with a conventional intelligent charging pile. Traditional intelligent charging stake mainly uses temperature sensor to monitor the temperature of charging stake to use simple threshold value to judge and carry out early warning suggestion. Conventional smart charging piles also use fixed charging power and time for charging and use manual means for settlement and payment of charging fees.

The results of the comparative experiments are shown in the following table:

From the table, the invention is superior to the traditional system in terms of fault recognition accuracy, efficiency, early warning prompt timeliness, detail, friendliness and convenience, and has obvious advantages and innovations. In addition, the invention is superior to the traditional system in the aspects of the charging power and time optimization rate, and the charging fee settlement and payment efficiency, and the invention has high intelligence and convenience.

Example 3: a intelligent electric pile that fills for electric bicycle

In this embodiment, the present invention is applied to an intelligent charging pile for an electric bicycle, which can automatically allocate and schedule resources of the charging pile according to the remaining capacity of the electric bicycle and the reservation time of a user, and realize reservation and cancellation of charging service.

The intelligent charging pile consists of a charging socket, an RFID tag, an infrared thermal imager, an acoustic wave sensor and a laser radar. The charging socket is used for being connected with a charging interface of the electric bicycle and providing electric energy for the charging interface. The RFID tag is used to identify and monitor the charging stake. The infrared thermal imager, the acoustic wave sensor and the laser radar are used for monitoring and measuring the temperature of the charging pile. The intelligent charging pile is also connected with an embedded module or a cloud platform and is communicated with the embedded module or the cloud platform through a wireless or wired communication line. The embedded module or cloud platform is provided with a blockchain system, an intelligent contract system and a deep learning model. The blockchain system is used to securely store, verify and transmit the data of the charging stake. The intelligent contract system is used for automatically processing and maintaining the data of the charging pile and realizing reservation and cancellation of charging service. The deep learning model is used for extracting, classifying and identifying the characteristics of the data of the charging pile, and optimizing the resource allocation and scheduling of the charging pile.

The working flow of the intelligent charging pile is as follows:

when an electric bicycle needs to reserve charging service, the electric bicycle sends a reservation request including the residual electric quantity and reservation time through a user terminal or a monitoring center. And the user terminal or the monitoring center sends the reservation request to the embedded module or the cloud platform.

The embedded module or the cloud platform receives the reservation request and verifies the validity and legality of the reservation request through the data verification contract. It then sends the reservation request to other nodes or systems via the data transfer contract.

The embedded module or the cloud platform also inputs all reservation requests into the deep learning model, performs feature extraction, classification and identification, performs matching and sequencing according to a pre-trained optimization algorithm, and outputs an optimal charging pile resource allocation and scheduling scheme.

And the embedded module or the cloud platform distributes and schedules corresponding charging pile resources through a data distribution contract according to an optimal scheme output by the deep learning model, and outputs reservation confirmation information.

In order to realize the data distribution contract, a distribution and scheduling algorithm based on a deep learning model is used, and the algorithm can match optimal charging pile resources for each user according to the reservation time, the residual electric quantity, the position, the state, the price and other factors of the charging pile of the user and output reservation confirmation information. The algorithm can be expressed by the following formula:

Where n is the number of users, m is the number of charging piles, c _ij is the total cost of the user i reserving charging piles j (including time cost, distance cost, price cost, etc.), x _ij is a binary variable indicating whether user i reserved charging piles j (1 indicates reservation, 0 indicates non-reservation), and λ is a regularization parameter for penalizing the user that no reservation or reservation of multiple charging piles is made.

The optimization problem may be solved by a deep learning model, for example using a multi-layer perceptron (MLP) to learn the functional relationship of c _ij and a softmax layer to output the probability distribution of x _ij. The model may be trained from a large amount of historical data and updated based on real-time data.

To implement the data payment contract, we can use a settlement and payment system based on blockchain technology, which can be sent to the corresponding paymate according to the user's choice and implement settlement and payment of prepaid fees. The system can be represented by the following flow:

when the user reserves the charging stake, selecting a payment mode (such as WeChat, payment treasures, bank cards and the like) and inputting verification information such as a payment password or fingerprint and the like;

The system encrypts the selection and verification information of the user and sends the encrypted selection and verification information to a corresponding payment platform, and requests to generate a payment certificate;

After the identity and balance of the user are verified by the payment platform, a payment certificate is generated, encrypted and returned to the system;

The system stores the payment certificate on the blockchain and takes the payment certificate as an input parameter of the intelligent contract;

When a user arrives at a charging pile and starts charging, the system automatically executes payment operation according to an intelligent contract and records a payment result on a blockchain;

when the user ends the charging and leaves the charging post, the system automatically performs a settlement operation according to the smart contract and records the settlement result on the blockchain.

The system can ensure the safety, transparency and traceability of payments while reducing intervention and cost of intermediaries

The embedded module or the cloud platform sends and displays reservation confirmation information to the user terminal or the monitoring center through the data display contract, and prompts the user to confirm or cancel the reservation.

The user terminal or the monitoring center receives and displays the reservation confirmation information and prompts the user to confirm or cancel the reservation. If the user confirms the reservation, the user terminal or the monitoring center sends the user selection to the embedded module or the cloud platform, and the embedded module or the cloud platform sends the user selection to the corresponding payment platform through the data payment contract, and the settlement and payment of the prepaid fee are realized. If the user cancels the reservation, the user terminal or the monitoring center sends the selection of the user to the embedded module or the cloud platform, and the embedded module or the cloud platform sends the selection of the user to the corresponding charging pile through the data cancellation contract and releases the resources of the charging pile.

When the user arrives at the reserved charging stake, he places his RFID card on the charging stake's RFID reader to verify his identity and balance. If the verification is successful, the charging pile unlocks the charging socket and prompts the user to insert the charging interface.

When a user inserts the charging interface, the charging pile starts to provide preset charging power and time for the user, and basic information and temperature data of the charging pile are sent to the embedded module or the cloud platform through the RFID tag.

The user terminal or the monitoring center receives and displays the early warning level and the emergency measures, and prompts the user to execute corresponding operations, such as cutting off the power supply of the charging pile, spraying water to extinguish fire, notifying fire and the like.

The user terminal or the monitoring center sends the user selection to the embedded module or the cloud platform, and the embedded module or the cloud platform sends the user selection to corresponding equipment or a system through a data maintenance contract, such as cutting off the power supply of the charging pile, spraying water to extinguish fire, notifying fire and the like.

When the charging pile is charged, the charging pile sends basic information and temperature data to the embedded module or the cloud platform through the RFID tag, and prompts a user to pull out a charging interface of the charging pile.

When a user pulls out the charging interface, the charging pile stops supplying electric energy and locks the charging socket. The embedded module or the cloud platform receives the data of the charging pile and verifies the validity and legality of the charging pile through a data verification contract. It then sends the charging stake data to other nodes or systems via a data transfer contract.

The intelligent charging pile can automatically allocate and schedule the resources of the charging pile according to the residual electric quantity of the electric bicycle and the reservation time of a user, and realize reservation and cancellation of charging service, thereby effectively improving the safety, reliability, efficiency and intelligence of a charging pile monitoring and early warning system.

In order to verify the effect of the present invention, a comparative experiment was performed in this example, and the present invention was compared with a conventional intelligent charging pile. Traditional intelligent charging stake mainly uses temperature sensor to monitor the temperature of charging stake to use simple threshold value to judge and carry out early warning suggestion. Traditional intelligent charging piles also use manual means to make reservations and cancellation of charging services, and resource allocation and scheduling.

The results of the comparative experiments are shown in the following table:

From the table, the invention is superior to the traditional system in terms of fault recognition accuracy, efficiency, early warning prompt timeliness, detail, friendliness and convenience, and has obvious advantages and innovations. In addition, the invention is superior to the traditional system in the aspects of accuracy and efficiency of reservation of the charging service, accuracy and efficiency of cancellation of the charging service and optimization rate of resource allocation and scheduling, and has high intelligence and flexibility.

Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.

Claims

1. The utility model provides a charging early warning method of charging stake, includes carrying out real-time supervision, data analysis and early warning suggestion to charging stake, its characterized in that: the real-time monitoring is used for carrying out high-definition monitoring and accurate temperature measurement on the charging pile and carrying out intelligent identification and remote monitoring on the charging pile; the data analysis comprises the steps of safely storing, verifying and transmitting the data of the charging pile, and extracting, classifying and identifying the characteristics of the data of the charging pile based on a deep learning algorithm; the early warning prompt is used for carrying out intelligent early warning of abnormal temperature rise, overtemperature or fire disaster according to the temperature data and the fault mode, and displaying early warning states and emergency measures through a user terminal or a monitoring center;

the abnormal temperature rise is identified by a temperature fluctuation score and a temperature convexity score:

；

wherein t _i,j represents the temperature of the jth sub-zone in the ith round of detection, The average temperature of all subareas in the ith round of detection is represented, and n1 represents the number of subareas; the temperature fluctuation score reflects the degree of change of the temperature in the charging pile area, and the larger the temperature fluctuation score is, the more unstable the temperature is;

The temperature convexity score cfs _i is calculated by the following formula:

；

wherein t _i,j represents the temperature of the jth sub-region in the ith round of detection, t _i represents the average temperature of all the sub-regions in the ith round of detection, and n ₁ represents the number of sub-regions; the temperature convexity score reflects the peak value or valley value degree of the temperature in the charging pile area, and the larger the temperature is, the less smooth the temperature is;

Determining a temperature detection score of each sub-region according to the temperature fluctuation score and the temperature convexity score, and determining a sub-region with abnormal temperature according to the temperature detection score;

The intelligent identification and remote monitoring of the charging pile are based on an RFID technology, and the method comprises the following steps:

(3) After receiving the response signal, the RFID reader-writer sends the data in the response signal to the embedded module or the cloud platform through the communication line of the Internet of things;

the safe storage, verification and transmission of the data of the charging pile are based on a blockchain technology and an intelligent contract technology, and the method comprises the following steps:

2. The charging early warning method of the charging pile according to claim 1, wherein the method comprises the following steps: the high-definition monitoring and accurate temperature measurement of the charging pile are based on an infrared thermal imaging technology, and the method comprises the following steps of:

3. The charging early warning method of the charging pile according to claim 1, wherein the method comprises the following steps: the real-time monitoring and accurate temperature measurement of the charging pile are further based on the acoustic wave sensor technology, and the method comprises the following steps:

4. The charging early warning method of the charging pile according to claim 3, wherein: the real-time monitoring and accurate temperature measurement of the charging pile are also based on a laser radar technology, and the method comprises the following steps:

(1) A laser radar transmitter transmits a laser signal to the charging pile;

5. The charging early warning method of the charging pile according to claim 1, wherein the method comprises the following steps: the deep learning algorithm-based feature extraction, classification and identification are carried out on the data of the charging pile, and the method comprises the following steps:

6. The charging early warning method of the charging pile according to claim 1, wherein the method comprises the following steps: the early warning prompt comprises intelligent early warning of abnormal temperature rise, overtemperature or fire disaster according to temperature data and fault modes, and early warning states and emergency measures are displayed through a user terminal or a monitoring center, and the method comprises the following steps:

(4) The user terminal or the monitoring center receives and displays the early warning state and the emergency measures, and executes corresponding power-off, water spraying fire extinguishing and fire extinguishing notification operations according to the selection of the user.