Communication security self-powered non-contact equipment diagnosis system and method based on AI pre-judgment

Abstract

The invention relates to the technical field of communication security and wireless energy supply, and particularly provides a communication security self-powered non-contact equipment diagnosis system and method based on AI pre-judgment. The method comprises the steps that a diagnosed device is safely awakened through NFC encryption authentication in a dormant state; the diagnosis equipment collects environment data, and transmission parameters are pre-judged through an AI model; according to the power supply state of the equipment, energy is supplied to the equipment through a wireless electric energy induction coil in an offline scene; establishing a satellite flash communication link and carrying out secondary encryption handshake to realize secure data interaction; energy supply power and communication parameters are optimized in real time based on an AI pre-judgment result, and the model is updated through feedback data. The system comprises a diagnosed device provided with a receiving coil, an encryption chip and a satellite flash module, and a diagnosis device provided with a transmitting coil, an NFC module, a satellite flash module and an AI pre-judgment module, and solves the safety problems that off-line device energy supply operation is tedious, communication is prone to being interfered, and illegal device awakening is caused. Efficient and safe diagnosis under the condition of no physical interface and non-contact is realized.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of data security and wireless communication and wireless energy transmission, in particular to a communication security self-powered non-contact device diagnosis system and method based on AI prediction. BACKGROUND

[0002] The existing device diagnosis technology mainly includes two types of mainstream schemes: one type is interface diagnosis, which realizes data interaction by connecting a diagnosis wire harness through a reserved physical interface such as a vehicle OBD physical interface, but the physical interface not only causes a security risk of core data leakage of the device, but also conflicts with the structure design requirement of waterproof and dustproof of the device; the other type is wireless diagnosis, which realizes non-contact interaction based on Bluetooth, cellular communication and the like, but needs to continuously broadcast electromagnetic waves, is not suitable for secret places, and still needs to be powered in a non-diagnosis mode, causing energy waste and being unable to adapt to energy-sensitive scenes.

[0003] There are contradictions between the physical interface and safety and structure requirements; the electromagnetic leakage and energy waste problems of wireless diagnosis; especially in the off-line diagnosis environment of scenes such as accident identification, complex wiring is needed for power supply, the operation is tedious and inefficient, the communication safety and efficiency are difficult to balance, and there is a lack of a cooperative mechanism of device legality authentication and data encryption; and the fixed technical configuration cannot adapt to dynamic scenes, resulting in insufficient power supply efficiency and anti-interference ability.

[0004] In the prior art, CN115884144B focuses on abnormal self-recovery of an NFC chip and does not involve power supply adaptation and device diagnosis; CN120730445A pays attention to low-power-consumption communication of a wireless BMS and lacks safety authentication and off-line power supply; and CN120916142A focuses on safe communication of a wireless BMS, but lacks dynamic power supply and scene adaptation.

[0005] Therefore, it is urgent to enhance scene adaptation by AI prediction technology on the basis of self-powered, non-contact and off-line diagnosis, solve the above-mentioned compound technical contradictions and improve the universality of the scheme. SUMMARY

[0006] The main purpose of the application is to provide a communication security self-powered non-contact device diagnosis system and method based on AI prediction, which solves the problems of insufficient dynamic adaptation of power supply efficiency and communication distance and safety of illegal device wake-up.

[0007] To solve the above technical problems, the technical scheme adopted by the application is as follows: a communication security self-powered non-contact device diagnosis method based on AI prediction comprises the following steps: S1, the device to be diagnosed is in a power-off state, only the NFC authentication function is reserved; the diagnosis device is started and is close to the device to be diagnosed and triggers the NFC encryption authentication; S2. When the diagnostic device detects that the NFC signal strength reaches the preset authentication value, it sends a pairing key to start authentication. If the detected device verifies the signal strength, it proceeds to the next step. S3. The detection equipment collects environmental data and uses AI to generate predictive transmission parameters based on the environmental data. S4. The diagnostic equipment uses wireless charging technology to power the device under test based on the transmission parameters predicted by AI. S5. Start the StarFlash communication link, configure anti-interference parameters according to the transmission parameters predicted by AI, and perform a second encrypted handshake. After passing the handshake, start bidirectional data interaction. S6. Based on AI prediction results, optimize wireless power supply parameters and adjust starlight communication parameters in real time; update model parameters by collecting real-time data through feedback parameters and synchronize them to the AI ​​prediction model. S7. After the data interaction is completed, the power supply stops in the offline scenario, the diagnostic module on the diagnosed end returns to sleep mode, and the StarSpark link is disconnected; in the online scenario, the StarSpark link is directly disconnected.

[0008] In the preferred embodiment, in step S1, the NFC encryption authentication adopts a combination mechanism of hash verification and AES encryption. The physical encryption chip of the device being diagnosed pre-stores the pairing key and the device's unique identifier. After the diagnostic device is started, it automatically loads the pre-stored key information.

[0009] In the preferred embodiment, in step S2, the pairing key verification logic is as follows: the diagnostic device sends an authentication frame containing the key hash value, which is decrypted by the device being diagnosed and compared with the locally pre-stored key hash value. If the verification matches, the diagnostic module is awakened; otherwise, the authentication fails.

[0010] In the preferred embodiment, in step S3, the environmental data includes the NFC received signal strength value, electromagnetic interference intensity, and real-time current and voltage of the device being diagnosed; the AI ​​adopts a lightweight long short-term memory network model, and the output predicted transmission parameters include energy consumption demand value, interference intensity level, and star-flash communication anti-interference configuration suggestions.

[0011] In the preferred embodiment, step S4 further includes: the wireless power supply power is related to the communication distance, and the output power is greater when the communication distance is longer. During the power supply process, the power supply voltage feedback of the device under diagnosis is continuously collected, and the power is dynamically adjusted based on the AI ​​prediction result when the voltage deviation is greater than the preset deviation value.

[0012] In the preferred embodiment, step S5 includes configuring anti-interference parameters such as the bandwidth of the StarFlash communication channel and the length of the checksum; the temporary session key generated during the secondary encryption handshake multiplexes the NFC authentication stage and is encrypted using the cryptographic block linking mode.

[0013] In the preferred embodiment, step S6 further includes: the real-time optimization logic is: when the AI ​​predicts that the interference intensity is increasing, the length of the Star Flash communication check code is increased and the transmission rate is reduced.

[0014] A communication security self-powered non-contact device diagnostic system based on AI prediction includes a device under test and a detection device. By introducing AI prediction technology, it solves the problems of insufficient dynamic adaptation between power supply efficiency and communication distance, and the security risks of device wake-up.

[0015] The device under test includes a receiving induction coil, a physical encryption chip, a star-flash communication module, and a diagnostic module; The testing equipment includes a transmitting induction coil, an NFC module, a StarFlash communication module, and an AI prediction module.

[0016] In the preferred embodiment, a receiving induction coil and a transmitting induction coil are used to generate a directional electromagnetic field to transmit and receive electrical energy.

[0017] In the preferred embodiment, the physical encryption chip pre-stores a pairing key, a unique device identifier, and encryption algorithm firmware for use in authenticating and testing the device.

[0018] In the preferred embodiment, the StarScan communication module is used to support the StarScan protocol and realize StarScan communication.

[0019] In the preferred embodiment, the diagnostic module uses a low-power MCU to collect the operating status parameters of the device being diagnosed.

[0020] In the preferred embodiment, the NFC module integrates encrypted authentication firmware for NFC authentication.

[0021] In the preferred embodiment, the AI ​​prediction module includes a scene parameter acquisition unit and a model calculation unit; the scene parameter acquisition unit is used to collect current and voltage during electromagnetic interference and wireless power transmission, and simultaneously collect multi-dimensional environmental data; the model calculation unit is used to predict and correct parameters based on environmental data.

[0022] This invention provides a diagnostic system and method for self-powered contactless devices with communication security based on AI prediction. It solves the problem of unauthorized device wake-up by using NFC encryption authentication and external wake-up-only functionality. AI prediction drives dynamic adaptation of power supply and communication parameters, and continuous optimization is achieved through closed-loop feedback. While retaining the core advantages of contactless, self-powered, and low-power consumption, it significantly improves power supply efficiency and scenario adaptability, adapting to the needs of multiple scenarios such as waterproof batteries, industrial controllers, and offline accident identification. Attached Figure Description

[0023] The present invention will be further described below with reference to the accompanying drawings and embodiments: Fig. 1 This is a flowchart of a diagnostic method for self-powered non-contact communication security devices based on AI prediction, according to the present invention. Fig. 2This is a schematic diagram of a self-powered non-contact device diagnostic system for communication security based on AI prediction, according to the present invention. In the diagram: Device under diagnosis 1, receiving induction coil 11, physical encryption chip 12, first star flash communication module 13, diagnostic module 14, first NFC module 15, diagnostic device 2, transmitting induction coil 21, second NFC module 22, second star flash communication module 23, AI prediction module 24. Detailed Implementation

[0024] like Figs. 1-2 As shown, a diagnostic system and method for self-powered non-contact communication security devices based on AI prediction are presented. Taking the offline accident identification scenario of waterproof batteries as a specific application scenario, the implementation process of the method and system is described in detail. All parameter values ​​are based on the preferred solutions of commercial components and documentation. In a preferred embodiment, a diagnostic method for self-powered contactless communication security devices based on AI prediction includes the following steps: S1. The device under test 1 is in a power-off state. After the diagnostic device 2 is started, it approaches the device under test 1 to the NFC signal detection range, driving the transmitting induction coil 21 to generate a directional electromagnetic field to power the device under test 1. The receiving induction coil 11 of the device under test 1 receives the electromagnetic field energy, which is rectified and then used to power the device under test 1. Through the external transistor of the physical encryption chip 12, power is only supplied to the physical encryption chip 12 and the NFC receiving circuit before the verification is passed, and the NFC encryption authentication process is triggered at the same time. The NFC encryption authentication adopts a combination mechanism of hash verification and AES encryption. The physical encryption chip 12 of the device under test 1 has a pre-stored pairing key and a unique device identifier. After the diagnostic device 2 is started, it automatically loads the pre-stored pairing key information from the local anti-tamper storage area to avoid the risk of plaintext transmission. S2. The second NFC module 22 of the diagnostic device 2 collects the NFC signal strength. When the detected signal strength is continuously greater than the preset minimum strength, it is determined that the preset authentication value has been reached, and an authentication frame containing the diagnostic device ID and the pairing key hash value is sent to the device under diagnosis 1. After receiving the authentication frame, the physical encryption chip 12 of the device under diagnosis 1 decrypts and extracts the hash value, and compares it with the locally pre-stored pairing key hash value. If the verification is successful, the physical encryption chip 12 outputs a start signal to the diagnostic module 14; If the verification fails, an authentication failure signal will be returned, and information on the illegal device will be reported, at which point the process will terminate. S3. Activate the scene parameter acquisition unit of diagnostic device 2 to collect NFC signal RSSI value, electromagnetic interference intensity, current of the device under diagnosis 1, and preset security level in real time. After the data collection is completed, the environmental data is input into the long short-term memory network model of the AI ​​prediction module 24 to generate the predicted transmission parameters: stable communication distance value, energy consumption requirement value, interference intensity level, and anti-interference configuration of star flash communication. S4. The power control circuit, based on the stable value of the communication distance predicted by AI, queries the preset distance-power mapping model to optimize power supply. During the power supply process, the diagnostic device 2 periodically reads the power supply voltage of the device under diagnosis 1. If the voltage deviation is greater than the preset deviation value, the AI ​​prediction module 24 dynamically adjusts the power supply to maintain voltage stability. S5. The second StarScan communication module 23 of diagnostic device 2 receives the anti-interference configuration suggestion predicted by AI, configures the channel bandwidth and checksum length, and then starts link scanning to establish a connection with the first StarScan communication module 13 of the device under diagnosis 1. It initiates a secondary encrypted handshake, reuses the AES temporary session key generated during the NFC authentication phase, and encrypts it in CBC mode. After the first StarScan communication module 13 of the device under diagnosis 1 decrypts and verifies the data, it returns a handshake response. After passing the verification, the diagnostic module 14 of the device under diagnosis 1 collects the battery operating parameters and uploads them through the StarScan link. After receiving the data, diagnostic device 2 completes bidirectional data interaction. S6. When the diagnostic device 2 detects a decrease in electromagnetic interference intensity, the first star-flash communication module 13 and the second star-flash communication module 23 switch the channel bandwidth to a higher frequency, reduce the check code length, and increase the transmission rate. When the battery power consumption current exceeds the peak preset value, an instruction is sent to the power control circuit to increase the power supply. At the same time, the input power and battery absorbed power, the ratio of the number of successfully transmitted and received star-flash data packets to the total number of transmitted packets, the total power consumption of the transmitting induction coil 21, and the bit error rate are collected in real time. When the deviation rate of any indicator is greater than the preset deviation value, the distance-power mapping model parameters and the weight of the long short-term memory network model are automatically updated and synchronized to the AI ​​prediction module 24. S7. When the diagnostic device 2 detects that the star-flash link has no data transmission for a preset duration, it determines that the data interaction is complete and the interaction is terminated. After the diagnostic module 14 in the device being diagnosed saves the diagnostic data to the local machine, it shuts down the first star-flash communication module 13 and the physical encryption chip 12, enters the sleep mode, shuts down the power supply of the transmitting induction coil 21, disconnects the star-flash link, and the process is terminated.

[0025] In the preferred embodiment, the NFC encryption authentication in step S1 also includes a replay attack prevention mechanism: The authentication frame sent by diagnostic device 2 contains a 4-byte incrementing sequence number. The sequence number is incremented by 1 with each authentication. The physical encryption chip 12 of the device being diagnosed stores the sequence number of the previous authentication. If the received sequence number is less than or equal to the stored value, it is determined to be a duplicate attack, and the authentication is directly canceled without hash verification.

[0026] In the preferred scheme, the training process of the long short-term memory network model in step S3 adopts the TensorFlow Lite framework. The training dataset includes distance values ​​in the range of 5-15cm, interference values ​​from -40dBm to 0dBm, energy consumption values ​​of 10-50mA, multiple sets of basic data, and superimposed interference fluctuation data. During training, batch gradient descent is used for multiple iterations.

[0027] In the preferred embodiment, the distance-power mapping model parameters in step S4 are determined by testing the power supply at distances of 5cm, 8cm, 12cm, and 15cm in an unobstructed environment. Based on this experimental data, the mapping relationship is determined to maintain the power supply efficiency at each distance to meet the operational requirements.

[0028] A communication security self-powered contactless device diagnostic system based on AI prediction includes a device under test 1 (battery) and a diagnostic device 2 (portable diagnostic instrument). The device under test 1 and the diagnostic device 2 achieve contactless interaction through NFC encryption authentication, wireless power transmission, and Star Flash communication. The system dynamically adapts power supply and communication parameters through an AI prediction module 24 to solve problems such as insufficient power supply efficiency and distance adaptation, and unauthorized wake-up.

[0029] The device under diagnosis 1 includes a receiving induction coil 11, a physical encryption chip 12, a first star flash communication module 13, a diagnostic module 14, and a first NFC module 15; The diagnostic device 2 includes a transmitting induction coil 21, a second NFC module 22, a second Star Flash communication module 23, and an AI prediction module 24.

[0030] In the preferred embodiment, the receiving induction coil 11 of the device under test 1 and the transmitting induction coil 21 of the diagnostic device 2 are both multi-turn hollow coils, using oxygen-free copper wire, adapted to the mega-level wireless charging frequency band, connected to and processed by the original integrated battery power management module of the device, converting the received electromagnetic energy into electrical energy.

[0031] In the preferred embodiment, both the first NFC module 15 and the second NFC module 22 are NXPPN532 models, with a communication distance of less than 15cm. They integrate NFC encryption and authentication firmware, can pre-store the device ID hash value and pairing key of the device being detected, and connect to the AI ​​prediction module.

[0032] In the preferred embodiment, the diagnostic module 14 of the device being diagnosed 1 uses an STM32L476RG microcontroller unit, which retains only NFC wake-up capability. After wake-up, it collects battery parameters through the I2C interface.

[0033] In the preferred embodiment, the physical encryption chip 12 of the device under diagnosis 1 is the AT88SC0104C model, which integrates SHA-256 and AES-128 hardware acceleration units. The tamper-proof storage area is pre-stored with a 256-bit pairing key and a 16-bit unique device ID, and only supports reading the storage area data via NFC commands. External pins cannot be directly accessed.

[0034] In the preferred embodiment, the first StarScan communication module 13 and the second StarScan communication module 23 of the device under diagnosis 1 and the diagnostic device 2 adopt the HX-STAR100 model, support the StarScan 1.0 protocol, integrate an AES-128 hardware encryption unit, and are connected to the diagnostic module 14 through the SPI interface. They can automatically switch the channel bandwidth according to the interference intensity.

[0035] In the preferred embodiment, the AI ​​prediction module 24 of the diagnostic device 2 includes a scene parameter acquisition unit, a model calculation unit, and a parameter output unit: the scene parameter acquisition unit integrates an AD8318 electromagnetic interference sensor, an ACS712 current sensor, and an INA219 voltage sensor, and the sensor data is transmitted through an I2C interface; the model calculation unit uses an STM32H743VI AI chip, which can complete word long short-term memory network model inference within 10ms; the parameter output unit outputs power supply commands to the power control circuit and anti-interference configuration commands to the star flash module through a GPIO interface.

[0036] In the preferred embodiment, the diagnostic device 2 also includes a closed-loop feedback module, which is integrated into the AI ​​prediction module 24. It periodically collects power supply efficiency, communication success rate, energy consumption, and bit error rate, and communicates with the host computer through the UART interface to display the data in real time. When the deviation rate of a certain indicator is greater than the preset deviation value, the AI ​​model update interface is automatically called to adjust the distance-to-power mapping table and the weights of the long short-term memory network model.

[0037] The above embodiments are merely preferred technical solutions of the present invention and should not be regarded as limitations on the present invention. The scope of protection of the present invention should be the technical solution described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. A diagnostic method for self-powered contactless communication security devices based on AI prediction, characterized in that, Includes the following steps: S1. The device under diagnosis is powered off, with only NFC authentication function remaining; after the diagnostic device is powered on, it is brought close to the device under diagnosis to trigger NFC encrypted authentication. S2. When the diagnostic device detects that the NFC signal strength reaches the preset authentication value, it sends a pairing key to start authentication. If the detected device verifies the signal strength, it wakes up the device. S3. The detection equipment collects environmental data and uses AI to generate predictive transmission parameters based on the environmental data. S4. The diagnostic device reads the status of the device under test via NFC commands and supplies power to the device under test using a wireless power induction coil based on the transmission parameters predicted by AI. S5. Start the StarFlash communication link, configure anti-interference parameters according to the transmission parameters predicted by AI, and perform a second encrypted handshake. After passing the handshake, start bidirectional data interaction. S6. Based on AI prediction results, optimize wireless power supply parameters and adjust starlight communication parameters in real time; update model parameters by collecting real-time data through feedback parameters and synchronize them to the AI ​​prediction model. S7. After the data interaction is completed, the power supply stops in the offline scenario, the diagnostic module on the diagnosed end returns to sleep mode, and the StarSpark link is disconnected; in the online scenario, the StarSpark link is directly disconnected.

2. The diagnostic method for self-powered contactless communication security devices based on AI prediction according to claim 1, characterized in that, In step S1, NFC encryption authentication uses a combination of hash verification and AES encryption mechanism, and the device being diagnosed has a pre-stored pairing key and a unique device identifier.

3. The diagnostic method for self-powered contactless communication security devices based on AI prediction according to claim 1, characterized in that, In step S2, the diagnostic device sends an authentication frame containing a key hash value. After being decrypted by the diagnostic device, the frame is compared with the local key hash value. If the verification matches, the diagnostic module is activated; otherwise, the authentication fails.

4. The diagnostic method for self-powered contactless communication security devices based on AI prediction as described in claim 1, characterized in that... In step S3, the environmental data includes the NFC received signal strength value, electromagnetic interference intensity, and real-time current and voltage of the device under diagnosis; the AI ​​adopts a lightweight long short-term memory network model, and the output predicted transmission parameters include energy consumption demand value, interference intensity level, and star-flash communication anti-interference configuration suggestions.

5. The diagnostic method for self-powered contactless communication security devices based on AI prediction according to claim 1, characterized in that, Step S4 also includes: S41. The wireless power supply power is adjusted in real time according to the communication distance. The output power is greater when the communication distance is longer. During the power supply process, the power supply voltage feedback of the device under diagnosis is continuously collected. When the voltage deviation is greater than the preset deviation value, the power is dynamically adjusted based on the AI ​​prediction result.

6. The diagnostic method for self-powered contactless communication security devices based on AI prediction according to claim 1, characterized in that, In step S5, the anti-interference parameter configuration includes the bandwidth of the Star Flash communication channel and the length of the check code; the temporary session key generated during the secondary encryption handshake multiplexes the NFC authentication stage and is encrypted using the cryptographic block linking mode.

7. A diagnostic system for self-powered, contactless communication security devices based on AI prediction, characterized in that, Includes the device being tested and the testing equipment, wherein: The device under test includes a receiving induction coil, a physical encryption chip, a star-flash communication module, and a diagnostic module; The testing equipment includes a transmitting induction coil, an NFC module, a StarFlash communication module, and an AI prediction module.

8. The AI-based predictive self-powered contactless device diagnostic system for communication security according to claim 7, characterized in that, The receiving induction coil and the transmitting induction coil are used to generate a directional electromagnetic field, which transmits electrical energy from the detection equipment to the equipment being detected. The physical encryption chip has a pre-stored pairing key, a unique device identifier, and encryption algorithm firmware, which is used for authenticating and testing devices.

9. A diagnostic system for self-powered, contactless communication security devices based on AI prediction, as described in claim 7, is characterized in that... The diagnostic module uses a low-power MCU to collect the operating status parameters of the device being diagnosed. The NFC module integrates encrypted authentication firmware for NFC authentication.

10. A diagnostic system for self-powered, contactless communication security devices based on AI prediction, as described in claim 7, characterized in that, The AI ​​prediction module includes a scene parameter acquisition unit and a model computation unit, wherein: The scene parameter acquisition unit is used to collect current and voltage during electromagnetic interference and wireless power transmission, and simultaneously collect multi-dimensional environmental data. The model computation unit is used to predict and correct parameters based on environmental data.

Images