An intelligent edge computing gateway and system with an asymmetric multi-processor architecture

Abstract

The present invention relates to the field of gateway technology, and discloses an intelligent edge computing gateway and system with an asymmetric multi-processor architecture, including a cloud control center, a cloud control center connection layer, a variable logic matrix, an offline execution management unit, an execution port and a data source; encrypted communication between the cloud control center and the cloud control center connection layer; a two-way signal connection between the output end of the cloud control center and the input end of the cloud control center connection layer. The intelligent edge computing gateway and system with an asymmetric multi-processor architecture realizes the separation of low-level hardware control functions and high-level logic control operation functions by configuring two processor units with different performance specifications on an integrated circuit PCB board, so as to achieve the edge computing processing functions and performance indicators that previously required two independent hardware to work together on one hardware carrier.

Description

Technical Field

[0001] The present invention relates to the field of gateway technology, and specifically to an intelligent edge computing gateway and system with an asymmetric multi-processor architecture. Background Art

[0002] An edge gateway is a gateway deployed at the edge of the network. It connects the physical and digital worlds through network connection, protocol conversion and other functions, and provides lightweight connection management, real-time data analysis and application management functions.

[0003] With the development of the Internet of Things, the application scenarios of the Internet of Things have higher and higher requirements on the computing power and real-time performance of the Internet of Things gateways. In particular, the application of edge computing has become increasingly important in industrial scenarios. In addition to the traditional physical interface connection with the terminal device, data communication protocol parsing and data transmission, the work content that the edge computing gateway device needs to handle also requires the addition of data caching, primary data combing, abnormal data alarm, data transmission quality monitoring, rapid event response, weak network and no network environment processing, and AI model deployment and application. Rich and complex applications have put forward higher requirements on the software and hardware system design of the edge computing gateway. If the ordinary computing server hardware architecture is directly adopted, it cannot meet the application requirements of the Internet of Things edge gateway in terms of usage cost, power consumption, volume and other indicators. If the conventional single-chip microcomputer architecture is adopted, it cannot meet the computing power requirements of advanced applications. Therefore, the present invention proposes an innovative asymmetric multi-processing intelligent edge computing gateway device with an architecture for realizing a method Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides an intelligent edge computing gateway and system with an asymmetric multi-processor architecture, which solves the problem that the cost, power consumption, volume and other indicators cannot meet the application requirements of the IoT edge gateway, and if the conventional single-chip microcomputer architecture is adopted, the computing power requirements of advanced applications cannot be met.

[0005] The present invention provides the following technical solutions: an intelligent edge computing gateway and system with an asymmetric multi-processor architecture, including a cloud control center, a cloud control center connection layer, a variable logic matrix, an offline execution management unit, an execution port and a data source;

[0006] Encrypted communication between the cloud control center and the cloud control center connection layer;

[0007] The output end of the cloud control center is bidirectionally signal-connected to the input end of the cloud control center connection layer, the output end of the cloud control center connection layer is bidirectionally signal-connected to the input end of the variable logic matrix, and the output end of the variable logic matrix is ​​bidirectionally signal-connected to the input end of the offline execution management unit, the input end of the execution port, and the input end of the data source, respectively.

[0008] Preferably, the cloud control center connection layer is used for data encryption, remote firmware upgrade, remote logic data loading, remote user code loading, identity authentication, remote management task loading, data transmission and device operation status transmission;

[0009] (1) Data encryption

[0010] Provide asymmetric key encryption functions such as ECC and RSA, dynamically generate symmetric exchange keys, and encrypt all data interacting with the control center;

[0011] (2) Remote firmware loading and upgrading

[0012] Upgrade the device firmware. During this process, the device will use the built-in encryption chip and dedicated certificate authentication encryption, and calculate the firmware hash value to ensure the integrity and security of the firmware;

[0013] The device has a firmware disaster recovery function. Even if the device is disconnected or even loses power during the upgrade process, the system will enable backup firmware recovery so that it can be upgraded to the latest firmware again.

[0014] (3) Remote logical data loading

[0015] Load data that has been compiled into logical data by the user or data center;

[0016] (4) Remote user code loading

[0017] Load the script program code submitted by the user in the control center;

[0018] (5) Identity Authentication

[0019] Used for mutual identity authentication between devices, users and control center servers. ECDSA signature is used to confirm that the identities of devices and servers are trustworthy.

[0020] (6) Remote management task loading

[0021] Load the scheduled or periodic tasks preset by the user or the control center to complete tasks such as inspection and scheduled self-inspection;

[0022] (7) Return data and return device operating status

[0023] Connect to the control center, send and receive instructions, and feedback execution result data and equipment operating status, historical operation records and alarm records.

[0024] Preferably, the variable logic matrix includes a logic abstraction layer, an interaction matrix, a logic execution module, a preset function module and a logic database;

[0025] The output end of the logic abstraction layer is bidirectionally connected to the input end of the interaction matrix, and the output end of the interaction matrix is ​​bidirectionally connected to the input end of the logic execution module, the input end of the preset function module and the input end of the logic database.

[0026] Preferably, the logic abstraction layer is used to implement user code parsing, control instruction parsing, physical data abstraction, output data reverse parsing and platform data parsing;

[0027] The user code parsing is to parse the user script code and abstract it into corresponding logical bytecode;

[0028] The control instruction parsing is to parse the control center instruction code and abstract it into corresponding logical data;

[0029] The physical data abstraction quantifies and abstracts the data input by the physical device into corresponding logic;

[0030] The output data is reversely parsed to translate the executed logic results into corresponding physical control quantities or standardized data;

[0031] The platform data analysis and data analysis provided by other data platforms abstract the corresponding logical nodes.

[0032] Preferably, the interactive matrix is ​​used to implement matrix operations, logic segmentation, logic synthesis, logic matrix intervention and data call;

[0033] The matrix operation provides logic matrix computing capability, and the logical computing capability of computing AND, NOT, OR, etc.;

[0034] The logic segmentation segments complex logic data into simple logic that can be quickly executed by the system;

[0035] The logic synthesis merges repeated logics to reduce storage space;

[0036] The logic matrix intervention replaces non-standard logic (logical operations not provided by the system will be defined as non-standard logic) with standardized logic intervention, and reorganizes the logic model with logic function modules (non-standard logic will be decomposed or combined into one or more standard logics according to mathematical theory);

[0037] The data call database and preset function models simplify user programming.

[0038] Preferably, the logic execution module includes ordered execution, thread execution, planned execution, execution result feedback, priority processing and termination execution;

[0039] The ordered execution executes the process in the logical data order (the calculation results are related to each other for ordered execution, the calculation results are not related and can be executed in parallel, and the calculation results are time-limited for planned execution. The order is determined by the user program execution needs);

[0040] The thread execution executes multiple processes in parallel in a multi-threaded manner;

[0041] Plan execution: Execute the process on time according to the predetermined plan;

[0042] Feedback results: provide the execution results of the process to other processes, or feed them back to the abstract layer as data output; priority processing: calculate the priority of each process and optimize process execution;

[0043] Terminate execution. When an unexpected exception occurs in the system, the abnormal process can be safely terminated, all resources of the process can be recovered, and memory leaks can be prevented.

[0044] Preferably, the offline execution management unit includes an offline execution unit, an offline strategy management unit, a planned task management unit, an offline operation history data management unit, an offline alarm processing unit and a power-off data protection unit;

[0045] The offline execution unit, when the system is detected to be offline, takes over the system and schedules and coordinates the execution process of the system;

[0046] The offline strategy management is to provide the execution strategy of the system in the offline state to maintain the optimal operation state;

[0047] The planned task management unit can continuously execute the preset task process when offline;

[0048] The offline operation history data management unit records all data, status, and execution results that occur during the offline process, and resubmits the history data after going online to maintain data continuity;

[0049] The offline alarm processing unit, when an alarm occurs while offline, will take corresponding actions according to a predetermined strategy to ensure the safety of the system and protect the control device;

[0050] The power-off data protection unit provides energy from a backup battery to maintain important data in a specific area when the system loses power. When the power is on, the data is rewritten to the memory or sent to the control center, thereby retaining the last state before the power failure and effectively improving the accuracy of system fault analysis.

[0051] Preferably, the encrypted communication comprises the following steps:

[0052] S801, the user logs into the system through the identity key chip, and the hardware true random number generator generates a dynamic key and transmits the information together with the identity key chip information to the connection layer of the cloud control center;

[0053] S802, the cloud control center connection layer dynamically generates a symmetric exchange key through asymmetric key encryption functions such as ECC and RSA of the data encryption module, and transmits the information to the cloud control center;

[0054] S803, the cloud control center connection layer receives feedback from the cloud control center and performs AES encryption or decryption processing based on the symmetrical exchange of keys according to the content of the information;

[0055] S804 The information processed by AES encryption or decryption enters the logic system and enters the shadow storage through the firmware to generate plaintext data.

[0056] Preferably, the logic model database is used to store pre-compiled logic data in the device and logic data function modules downloaded remotely.

[0057] The preset function module is used to include a variety of integrated functions, algorithms such as search calculation, format conversion, hash calculation, etc.

[0058] An intelligent edge computing gateway with an asymmetric multi-processor architecture, comprising an integrated circuit PCB board, and the following structures integrated on the integrated circuit PCB board: a main memory, a link arbitrator, a dual-port shared memory, an embedded processor and a microcontroller;

[0059] The main memory includes interpretation support for multiple scripting language programs, implements variable embedded user code, and is used to store data and programs and execute programs;

[0060] The link arbiter is used to coordinate access when the embedded processor and the microcontroller simultaneously access the same address of the dual-port shared memory, ensure the access result is correct, respond to interrupt requests through a priority mechanism, and select the working order of the embedded processor and the microcontroller according to the preset priority;

[0061] Dual-port shared memory is used for data exchange between embedded processor and microcontroller;

[0062] The embedded processor is used to parse the script language and transmit it to the microcontroller to complete the cloud control center connection layer. It has an encryption chip, hardware true random number generator, shadow storage, encryption authentication certificate, and firmware disaster recovery function to achieve cloud connection and encrypted communication.

[0063] The microcontroller is used to receive and execute the script language parsed by the embedded processor, and can cooperate with the embedded processor to complete the variable logic matrix control, connect with the physical device, realize the execution of the driving physical execution device, realize the offline execution management of the physical device, and use the battery for the internal device;

[0064] The main memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute a system of an intelligent edge computing gateway with an asymmetric multi-processor architecture.

[0065] Information processing flow description:

[0066] 1. The intelligent edge computing gateway is connected to the cloud control center (the control center platform can be deployed on a public cloud or a private cloud) through wired LAN Ethernet or wireless WIFI, 4G / 5G, etc.;

[0067] 2. The edge computing gateway conducts encrypted communication with the cloud to obtain operation instructions, and then enters the system's variable logic matrix (the variable logic matrix is ​​used to distribute the logical results of program execution to different hardware port outputs) for calculation and processing. The variable logic matrix can change the logical combination of control instructions, data, and execution systems, thereby changing the execution process of the system. The logic matrix can be changed remotely or locally on site.

[0068] For example: Users can remotely or instantly change the program execution results to different execution agencies according to their usage requirements. This change can be caused by user intervention, external input changes, or program execution process results.

[0069] 3. The data source interface is used to connect various physical devices or other software systems (other software systems refer to software developed by a third party that obtains data through a certain protocol) to collect relevant physical quantity data.

[0070] 4. The execution port is used to drive various physical execution devices to execute (the program execution results are output to various execution devices and must meet the input requirements of the device, so the digital results need to be converted into physical level output.) The corresponding logical commands.

[0071] 5. The offline execution management module is used to ensure normal operation of the edge when the connection between the device and the cloud control center is disconnected.

[0072] Compared with the prior art, the present invention provides an intelligent edge computing gateway and system with an asymmetric multi-processor architecture, which has the following beneficial effects:

[0073] (1) The intelligent edge computing gateway and system with an asymmetric multi-processor architecture realizes the separation of low-level hardware control functions and high-level logic control operation functions by configuring two processor units with different performance specifications on an integrated circuit PCB board, thereby achieving edge computing processing functions and performance indicators that previously required two independent hardware to work together on a single hardware carrier.

[0074] (2) The intelligent edge computing gateway and system with an asymmetric multi-processor architecture provides a solution that meets the application requirements of the IoT edge gateway in terms of cost, power consumption, volume, etc., while also meeting the computing power requirements of advanced applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 It is a schematic block diagram of the system logic structure of the present invention;

[0076] Figure 2 The overall design framework diagram of the hardware of the present invention;

[0077] Figure 3 This is a schematic block diagram of the encrypted communication logic structure of the present invention. DETAILED DESCRIPTION

[0078] See also Figure 1-3 , an intelligent edge computing gateway and system with an asymmetric multi-processor architecture, an intelligent edge computing gateway and system with an asymmetric multi-processor architecture, including a cloud control center, a cloud control center connection layer, a variable logic matrix, an offline execution management unit, an execution port and a data source;

[0079] Encrypted communication between the cloud control center and the cloud control center connection layer;

[0080] The output end of the cloud control center is bidirectionally signal-connected to the input end of the cloud control center connection layer, the output end of the cloud control center connection layer is bidirectionally signal-connected to the input end of the variable logic matrix, and the output end of the variable logic matrix is ​​bidirectionally signal-connected to the input end of the offline execution management unit, the input end of the execution port, and the input end of the data source, respectively.

[0081] Preferably, the cloud control center connection layer is used for data encryption, remote firmware upgrade, remote logic data loading, remote user code loading, identity authentication, remote management task loading, data transmission and device operation status transmission;

[0082] (1) Data encryption

[0083] Provide asymmetric key encryption functions such as ECC and RSA, dynamically generate symmetric exchange keys, and encrypt all data interacting with the control center;

[0084] (2) Remote firmware loading and upgrading

[0085] Upgrade the device firmware. During this process, the device will use the built-in encryption chip and dedicated certificate authentication encryption, and calculate the firmware hash value to ensure the integrity and security of the firmware;

[0086] The device has a firmware disaster recovery function. Even if the device is disconnected or even loses power during the upgrade process, the system will enable backup firmware recovery so that it can be upgraded to the latest firmware again.

[0087] (3) Remote logical data loading

[0088] Load data that has been compiled into logical data by the user or data center;

[0089] (4) Remote user code loading

[0090] Load the script program code submitted by the user in the control center;

[0091] (5) Identity Authentication

[0092] Used for mutual identity authentication between devices, users and control center servers. ECDSA signature is used to confirm that the identities of devices and servers are trustworthy.

[0093] (6) Remote management task loading

[0094] Load the scheduled or periodic tasks preset by the user or the control center to complete tasks such as inspection and scheduled self-inspection;

[0095] (7) Return data and return device operating status

[0096] Connect to the control center, send and receive instructions, and feedback execution result data and equipment operating status, historical operation records and alarm records.

[0097] Preferably, the variable logic matrix includes a logic abstraction layer, an interaction matrix, a logic execution module, a preset function module and a logic database;

[0098] The output end of the logic abstraction layer is bidirectionally connected to the input end of the interaction matrix, and the output end of the interaction matrix is ​​bidirectionally connected to the input end of the logic execution module, the input end of the preset function module and the input end of the logic database.

[0099] Preferably, the logic abstraction layer is used to implement user code parsing, control instruction parsing, physical data abstraction, output data reverse parsing and platform data parsing;

[0100] The user code parsing is to parse the user script code and abstract it into corresponding logical bytecode;

[0101] The control instruction parsing is to parse the control center instruction code and abstract it into corresponding logical data;

[0102] The physical data abstraction quantifies and abstracts the data input by the physical device into corresponding logic;

[0103] The output data is reversely parsed to translate the executed logic results into corresponding physical control quantities or standardized data;

[0104] The platform data analysis and data analysis provided by other data platforms abstract the corresponding logical nodes.

[0105] Preferably, the interactive matrix is ​​used to implement matrix operations, logic segmentation, logic synthesis, logic matrix intervention and data call;

[0106] The matrix operation provides logic matrix computing capability, and the logical computing capability of computing AND, NOT, OR, etc.;

[0107] The logic segmentation segments complex logic data into simple logic that can be quickly executed by the system;

[0108] The logic synthesis merges repeated logics to reduce storage space;

[0109] The logic matrix intervention replaces non-standard logic (logical operations not provided by the system will be defined as non-standard logic) with standardized logic intervention, and reorganizes the logic model with logic function modules (non-standard logic will be decomposed or combined into one or more standard logics according to mathematical theory);

[0110] The data call database and preset function models simplify user programming.

[0111] Preferably, the logic execution module includes ordered execution, thread execution, planned execution, execution result feedback, priority processing and termination execution;

[0112] The ordered execution executes the process in the logical data order (the calculation results are related to each other for ordered execution, the calculation results are not related and can be executed in parallel, and the calculation results are time-limited for planned execution. The order is determined by the user program execution needs);

[0113] The thread execution executes multiple processes in parallel in a multi-threaded manner;

[0114] Plan execution: Execute the process on time according to the predetermined plan;

[0115] Feedback results: provide the execution results of the process to other processes, or feed them back to the abstract layer as data output; priority processing: calculate the priority of each process and optimize process execution;

[0116] Terminate execution. When an unexpected exception occurs in the system, the abnormal process can be safely terminated, all resources of the process can be recovered, and memory leaks can be prevented.

[0117] Preferably, the offline execution management unit includes an offline execution unit, an offline strategy management unit, a planned task management unit, an offline operation history data management unit, an offline alarm processing unit and a power-off data protection unit;

[0118] The offline execution unit, when the system is detected to be offline, takes over the system and schedules and coordinates the execution process of the system;

[0119] The offline strategy management is to provide the execution strategy of the system in the offline state to maintain the optimal operation state;

[0120] The planned task management unit can continuously execute the preset task process when offline;

[0121] The offline operation history data management unit records all data, status, and execution results that occur during the offline process, and resubmits the history data after going online to maintain data continuity;

[0122] The offline alarm processing unit, when an alarm occurs while offline, will take corresponding actions according to a predetermined strategy to ensure the safety of the system and protect the control device;

[0123] The power-off data protection unit provides energy from a backup battery to maintain important data in a specific area when the system loses power. When the power is on, the data is rewritten to the memory or sent to the control center, thereby retaining the last state before the power failure and effectively improving the accuracy of system fault analysis.

[0124] Preferably, the encrypted communication comprises the following steps:

[0125] S801, the user logs into the system through the identity key chip, and the hardware true random number generator generates a dynamic key and transmits the information together with the identity key chip information to the connection layer of the cloud control center;

[0126] S802, the cloud control center connection layer dynamically generates a symmetric exchange key through asymmetric key encryption functions such as ECC and RSA of the data encryption module, and transmits the information to the cloud control center;

[0127] S803, the cloud control center connection layer receives feedback from the cloud control center and performs AES encryption or decryption processing based on the symmetrical exchange of keys according to the content of the information;

[0128] S804 The information processed by AES encryption or decryption enters the logic system and enters the shadow storage through the firmware to generate plaintext data.

[0129] Preferably, the logic model database is used to store pre-compiled logic data in the device and logic data function modules downloaded remotely.

[0130] The preset function module is used to include a variety of integrated functions, algorithms such as search calculation, format conversion, hash calculation, etc.

[0131] An intelligent edge computing gateway with an asymmetric multi-processor architecture, comprising a main memory, a link arbitrator, a dual-port shared memory, an embedded processor and a microcontroller;

[0132] The main memory includes interpretation support for multiple scripting language programs, implements variable embedded user code, and is used to store data and programs and execute programs;

[0133] The link arbiter is used to coordinate access when the embedded processor and the microcontroller simultaneously access the same address of the dual-port shared memory, ensure the access result is correct, respond to interrupt requests through a priority mechanism, and select the working order of the embedded processor and the microcontroller according to the preset priority;

[0134] Dual-port shared memory is used for data exchange between embedded processor and microcontroller;

[0135] The embedded processor is used to parse the script language and transmit it to the microcontroller to complete the cloud control center connection layer. It has an encryption chip, hardware true random number generator, shadow storage, encryption authentication certificate, and firmware disaster recovery function to achieve cloud connection and encrypted communication.

[0136] The microcontroller is used to receive and execute the script language parsed by the embedded processor, and can cooperate with the embedded processor to complete the variable logic matrix control, connect with the physical device, realize the execution of the driving physical execution device, realize the offline execution management of the physical device, and use the battery for the internal device;

[0137] The main memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute a system of an intelligent edge computing gateway with an asymmetric multi-processor architecture.

[0138] Information processing flow description:

[0139] 1. The intelligent edge computing gateway is connected to the cloud control center (the control center platform can be deployed on a public cloud or a private cloud) through wired LAN Ethernet or wireless WIFI, 4G / 5G, etc.;

[0140] 2. The edge computing gateway conducts encrypted communication with the cloud to obtain operation instructions, and then enters the system's variable logic matrix (the variable logic matrix is ​​used to distribute the logical results of program execution to different hardware port outputs) for calculation and processing. The variable logic matrix can change the logical combination of control instructions, data, and execution systems, thereby changing the execution process of the system. The logic matrix can be changed remotely or locally on site.

[0141] For example: Users can remotely or instantly change the program execution results to different execution agencies according to their usage requirements. This change can be caused by user intervention, external input changes, or program execution process results.

[0142] 3. The data source interface is used to connect various physical devices or other software systems (other software systems refer to software developed by a third party that obtains data through a certain protocol) to collect relevant physical quantity data.

[0143] 4. The execution port is used to drive various physical execution devices to execute (the program execution results are output to various execution devices and must meet the input requirements of the device, so the digital results need to be converted into physical level output.) The corresponding logical commands.

[0144] 5. The offline execution management module is used to ensure normal operation of the edge when the connection between the device and the cloud control center is disconnected.

[0145] Specific implementation method (give an example of a hardware physical structure implementation method, and the parameters of related components can be adjusted according to specific performance requirements)

[0146] 1. Core system processor

[0147] ARM v8 architecture cortex-A53 processor, 4 cores, 1-1.5Ghz main frequency, supports 1-2GRAM, runs customized arm-linux system, includes interpretation support for multiple scripting language programs. Realizes variable embedding of user code.

[0148] 2. Single chip control processor,

[0149] ARM v7 architecture cortex-M microprocessor, single core, 200Mhz main frequency, embedded 500KRAM and 1MFLASH. With 1.25DMIPS. Mainly provides various physical device control interfaces, and is also used to implement variable logic matrix operations.

[0150] 3. Dual-port memory system

[0151] Due to the use of a dual-processor structure, the memory system processing includes not only the traditional memory system, but also this dual-port memory control system to provide data sharing and fast switching between the two CPUs.

[0152] 4Data arbitration bridge module

[0153] It is used to determine which processor the data and instructions should be sent to for execution, and in which memory space they should be stored. It also determines whether the data should be shared or exchanged.

[0154] 5.FLASH, flash memory

[0155] Provides 32-128G sd-card storage with MMC interface for storing firmware, scheduled tasks, logical data, user code, and operating system.

[0156] 6. Wireless Network

[0157] Low-power narrowband IoT wireless interface.

[0158] 7. Network Interface

[0159] 1000M bandwidth Ethernet interface, 4G mobile data network interface.

[0160] 8. Physical device interface

[0161] Various digital, analog, switch, power and other physical device interfaces.

[0162] An example of data processing: When a user downloads a program to the edge gateway, the embedded processor will first analyze whether the program is executed by itself or by the microcontroller. If it is executed by itself, it will be stored in the main memory for execution. If it is executed by the microcontroller, it will first check whether it is a script language. If so, it will be parsed and converted into machine language for transmission to the microcontroller. Otherwise, it will be directly transmitted to the microcontroller. When executing the program, when data needs to be transferred from one CPU to another, it must be exchanged through the shared memory. The arbitrator is used to resolve conflicts during data exchange, resolve data confusion, and protect data integrity.

[0163] Example 1: The use method based on the patent of the present invention is explained in the application scenario of a central air-conditioning intelligent control system in a large shopping mall.

[0164] The central air-conditioning intelligent control system is mainly composed of three parts: the first layer is the data perception layer, the second layer is the Internet of Things control layer, and the third layer is the cloud-based intelligent management layer.

[0165] The data perception layer is mainly composed of various environmental and status sensors, including temperature and humidity sensors, environmental living thing sensors, motion and ambient light sensors, door and window opening and closing sensors, carbon dioxide and other gas sensors, pipeline flow sensors, etc.

[0166] The IoT control layer is mainly composed of a cold source control box, a wind cabinet control box, various water pumps and equipment switch controllers, an indoor temperature control panel, etc. The intelligent edge gateway of the patent of this invention is at this level and is installed and deployed in the above control box to obtain electrical energy through the power module in the control box.

[0167] The cloud-based intelligent management layer is mainly composed of a cloud-based software platform. Its main functions include the configurable display of the central air-conditioning system, presenting the data of all devices and all monitoring points in the system, configuring equipment parameters, configuring various intelligent control algorithm models, providing manual control of equipment, and various business management functions.

[0168] The intelligent edge computing gateway of the patent of this invention is installed in various control boxes of the Internet of Things control layer, and various sensors, cold source controllers, fan controllers, water pump switches, photoelectric switches, etc. of the data perception layer are connected to the physical interface of the intelligent edge computing gateway through RS485, RS232, USB, RJ45 and other communication lines.

[0169] The intelligent edge computing gateway is connected to the enterprise LAN through RJ45, fiber optic wired communication, or WIFI, 4G / 5G wireless communication, and then transferred to the Internet or directly connected to the Internet. It uses MQTT, HTTP, Websocket or private protocols to establish a link with the cloud intelligent management platform through the Internet.

[0170] The cloud-based intelligent management platform can send the intelligent control strategy of the central air conditioner to the intelligent edge computing gateway through data encryption through automatic triggering or manual sending. The embedded processor of the gateway mainly processes the control strategy from the cloud platform, parses and translates it into the real-time control instruction set of the microcontroller, and sends the instruction set to the microcontroller. The microcontroller uses a real-time operating system, and loads the real-time control instruction set of the intelligent control strategy sent by the embedded processor into the storage area for use by the offline execution management module.

[0171] The electrical signal sent by the sensor to the intelligent edge computing gateway through the signal line is converted into a digital signal by the microcontroller and transmitted to the embedded processor. The embedded processor transmits the data to the cloud platform. If the gateway and the cloud platform are disconnected from the network, the sensor data will be cached in the gateway storage and retransmitted to the cloud after the network is reconnected. The embedded processor is responsible for handling a series of processing tasks such as disconnection caching and breakpoint resumption to ensure stable and reliable communication with the cloud platform and data security.

[0172] The cloud platform sends control signals to operate the equipment, such as starting and stopping the water pump. The embedded processor sends the digital signal to the microcontroller, which converts the digital signal into an electrical signal and sends it to the equipment control switch.

[0173] The values ​​monitored by the sensor will first be processed by the microcontroller for strategy judgment. When the values ​​trigger the execution of a control strategy, the microcontroller will execute the action according to the established strategy and send the corresponding control electrical signal to the corresponding device switch to achieve real-time intelligent control. At the same time, the relevant status data will be pushed to the embedded processor for synchronous transmission to the cloud platform. Because the intelligent control strategy is deployed in the offline execution management module, intelligent decision-making and control can be achieved even when the network is disconnected from the cloud platform.

[0174] When the intelligent control strategy needs to be updated, you only need to define the control strategy on the cloud platform and then send it to the gateway. The embedded processor translates the new version of the strategy into a real-time control instruction set and sends it to the microcontroller. The microcontroller loads the new control instruction set, switches the instruction set in a safe and reliable manner, and notifies the embedded processor that the new strategy has been loaded successfully.

[0175] Through the intelligent edge computing gateway of the patent of this invention, the intelligent control strategy of the central air conditioner can be made available to management personnel in the form of configuration parameters. The intelligent control strategy obtained through the artificial intelligence large model algorithm can also be configured and loaded through this method. The high computing power and large storage of the embedded processor in the intelligent edge computing gateway can handle complex computing tasks well, while the microcontroller focuses on the real-time low-latency conversion processing of digital instructions and physical signals. While ensuring real-time intelligent control and offline intelligent control, the system platform has the ability to easily operate and dynamically change equipment intelligent management.

[0176] Example 2 illustrates the usage method based on the patent of the present invention in the application scenario of an intelligent control system for lighting fixtures in a large school.

[0177] The intelligent control system of lighting fixtures is mainly composed of three parts: the first layer is the data perception layer, the second layer is the Internet of Things control layer, and the third layer is the cloud-based intelligent management layer.

[0178] The data perception layer is mainly composed of various environmental and status sensors, including temperature and humidity sensors, environmental living thing sensors, motion and ambient light sensors, door and window opening and closing sensors, light intensity sensors, sound sensors, etc.

[0179] The IoT control layer is mainly composed of various types of lamps and equipment switch controllers, indoor lighting switch panels, etc. The intelligent edge gateway of the patent of this invention is at this level and is installed and deployed in the above control box to obtain electrical energy through the power module in the control box.

[0180] The cloud-based intelligent management layer is mainly composed of a cloud-based software platform. Its main functions include the configurable display of the intelligent lighting system, presenting the data of all devices and all monitoring points in the system, configuring device parameters, configuring various intelligent control algorithm models, providing manual control of equipment, and various business management functions.

[0181] The intelligent edge computing gateway of the patent of this invention is installed in various control boxes of the Internet of Things control layer, and various sensors of the data perception layer are connected to the physical interface of the intelligent edge computing gateway through communication lines such as RS485, RS232, USB, and RJ45.

[0182] The intelligent edge computing gateway is connected to the enterprise LAN through RJ45, fiber optic wired communication, or WIFI, 4G / 5G wireless communication, and then transferred to the Internet or directly connected to the Internet. It uses MQTT, HTTP, Websocket or private protocols to establish a link with the cloud intelligent management platform through the Internet.

[0183] The cloud-based intelligent management platform can send the intelligent control strategy of the central air conditioner to the intelligent edge computing gateway through data encryption through automatic triggering or manual sending. The embedded processor of the gateway mainly processes the control strategy from the cloud platform, parses and translates it into the real-time control instruction set of the microcontroller, and sends the instruction set to the microcontroller. The microcontroller uses a real-time operating system, and loads the real-time control instruction set of the intelligent control strategy sent by the embedded processor into the storage area for use by the offline execution management module.

[0184] The electrical signal sent by the sensor to the intelligent edge computing gateway through the signal line is converted into a digital signal by the microcontroller and transmitted to the embedded processor. The embedded processor transmits the data to the cloud platform. If the gateway and the cloud platform are disconnected from the network, the sensor data will be cached in the gateway storage and retransmitted to the cloud after the network is reconnected. The embedded processor is responsible for handling a series of processing tasks such as disconnection caching and breakpoint resumption to ensure stable and reliable communication with the cloud platform and data security.

[0185] The cloud platform sends control signals to operate the equipment, such as switching on and off lights in a specific classroom. The embedded processor sends the digital signal to the microcontroller, which converts the digital signal into an electrical signal and sends it to the equipment control switch.

[0186] The values ​​monitored by the sensor will first be processed by the microcontroller for strategy judgment. When the values ​​trigger the execution of a control strategy, the microcontroller will execute the action according to the established strategy and send the corresponding control electrical signal to the corresponding device switch to achieve real-time intelligent control. At the same time, the relevant status data will be pushed to the embedded processor for synchronous transmission to the cloud platform. Because the intelligent control strategy is deployed in the offline execution management module, intelligent decision-making and control can be achieved even when the network is disconnected from the cloud platform.

[0187] When the intelligent control strategy needs to be updated, you only need to define the control strategy on the cloud platform and then send it to the gateway. The embedded processor translates the new version of the strategy into a real-time control instruction set and sends it to the microcontroller. The microcontroller loads the new control instruction set, switches the instruction set in a safe and reliable manner, and notifies the embedded processor that the new strategy has been loaded successfully.

[0188] Through the intelligent edge computing gateway of the present invention, the intelligent control strategy of lighting registration can be made available to managers in the form of configuration parameters. The intelligent control strategy obtained by the artificial intelligence large model algorithm can also be configured and loaded through this method. The high computing power and large storage of the embedded processor in the intelligent edge computing gateway can handle complex computing tasks well, while the microcontroller focuses on the real-time low-latency conversion processing of digital instructions and physical signals. While ensuring real-time intelligent control and offline intelligent control, the system platform has the ability to manage equipment intelligently with convenient operation and dynamic changes.

[0189] To summarize, the intelligent edge computing gateway and system with an asymmetric multi-processor architecture realizes the separate computing control of low-level hardware control functions and high-level logic control operation functions by configuring two processor units with different performance specifications on an integrated circuit PCB board, so as to achieve edge computing processing functions and performance indicators that previously required two independent hardware to work together on one hardware carrier.

[0190] The intelligent edge computing gateway and system with an asymmetric multi-processor architecture provides a solution that meets the application requirements of the IoT edge gateway in terms of cost, power consumption, volume and other indicators, while also meeting the computing power requirements of advanced applications.

Claims

1. A system of an intelligent edge computing gateway with an asymmetric multi-processor architecture, characterized in that: It includes a cloud control center, a cloud control center connection layer, a variable logic matrix, an offline execution management unit, an execution port and a data source; Encrypted communication between the cloud control center and the cloud control center connection layer; The output end of the cloud control center is bidirectionally connected to the input end of the cloud control center connection layer, the output end of the cloud control center connection layer is bidirectionally connected to the input end of the variable logic matrix, and the output end of the variable logic matrix is ​​bidirectionally connected to the input end of the offline execution management unit, the input end of the execution port and the input end of the data source respectively; The cloud control center connection layer is used for data encryption, remote firmware upgrade, remote logic data loading, remote user code loading, identity authentication, remote management task loading, data transmission and device operation status transmission; The variable logic matrix includes a logic abstraction layer, an interaction matrix, a logic execution module, a preset function module and a logic database.

2. The system of the intelligent edge computing gateway with asymmetric multi-processor architecture according to claim 1, characterized in that: (1) Data encryption Provide asymmetric key encryption functions such as ECC and RSA, dynamically generate symmetric exchange keys, and encrypt all data interacting with the control center; (2) Remote firmware loading and upgrading Upgrade the device firmware. During this process, the device will use the built-in encryption chip and dedicated certificate authentication encryption, and calculate the firmware hash value to ensure the integrity and security of the firmware; The device has a firmware disaster recovery function. Even if the device is disconnected or even loses power during the upgrade process, the system will enable backup firmware recovery so that it can be upgraded to the latest firmware again. (3) Remote logical data loading Load data that has been compiled into logical data by the user or data center; (4) Remote user code loading Load the script program code submitted by the user in the control center; (5) Identity Authentication Used for mutual identity authentication between devices, users and control center servers. ECDSA signature is used to confirm that the identities of devices and servers are trustworthy. (6) Remote management task loading Load the scheduled or periodic tasks preset by the user or the control center to complete tasks such as inspection and scheduled self-inspection; (7) Return data and return device operating status Connect to the control center, send and receive instructions, and feedback execution result data and equipment operating status, historical operation records and alarm records.

3. The system of the intelligent edge computing gateway with asymmetric multi-processor architecture according to claim 1, characterized in that: The output end of the logic abstraction layer is bidirectionally connected to the input end of the interaction matrix, and the output end of the interaction matrix is ​​bidirectionally connected to the input end of the logic execution module, the input end of the preset function module and the input end of the logic database.

4. The system of the intelligent edge computing gateway with asymmetric multi-processor architecture according to claim 3 is characterized in that: The logic abstraction layer is used to implement user code parsing, control instruction parsing, physical data abstraction, output data reverse parsing and platform data parsing; The user code parsing is to parse the user script code and abstract it into corresponding logical bytecode; The control instruction parsing is to parse the control center instruction code and abstract it into corresponding logical data; The physical data abstraction quantifies and abstracts the data input by the physical device into corresponding logic; The output data is reversely parsed to translate the executed logic results into corresponding physical control quantities or standardized data; The platform data analysis and data analysis provided by other data platforms abstract the corresponding logical nodes.

5. The system of the intelligent edge computing gateway with asymmetric multi-processor architecture according to claim 3, characterized in that: The interactive matrix is ​​used to realize matrix operation, logic segmentation, logic synthesis, logic matrix intervention and data call; The matrix operation provides logic matrix computing capability, and the logical computing capability of computing AND, NOT, OR, etc.; The logic segmentation segments complex logic data into simple logic that can be quickly executed by the system; The logic synthesis merges repeated logics to reduce storage space; The logic matrix intervention replaces non-standard logic with standardized logic intervention and restructures the logic model with logic function modules; The data call database and preset function models simplify user programming.

6. The system of the intelligent edge computing gateway with asymmetric multi-processor architecture according to claim 3, characterized in that: The logic execution module includes ordered execution, thread execution, planned execution, execution result feedback, priority processing and termination execution; The ordered execution executes the process in a logical data sequence; The thread execution executes multiple processes in parallel in a multi-threaded manner; Plan execution: Execute the process on time according to the predetermined plan; Feedback results: provide the execution results of the process to other processes, or feed them back to the abstract layer as data output; Priority processing calculates the priority of each process and optimizes process execution; Terminate execution. When an unexpected exception occurs in the system, the abnormal process can be safely terminated and all resources of the process can be recovered.

7. The system of the intelligent edge computing gateway with asymmetric multi-processor architecture according to claim 1, characterized in that: The offline execution management unit includes an offline execution unit, an offline strategy management unit, a planned task management unit, an offline operation history data management unit, an offline alarm processing unit and a power-off data protection unit; The offline execution unit, when the system is detected to be offline, takes over the system and schedules and coordinates the execution process of the system; The offline strategy management is to provide the execution strategy of the system in the offline state to maintain the optimal operation state; The planned task management unit can continuously execute the preset task process when offline; The offline operation history data management unit records all data, status, and execution results that occur during the offline process, and resubmits the history data after going online to maintain data continuity; The offline alarm processing unit, when an alarm occurs while offline, will take corresponding actions according to a predetermined strategy to ensure the safety of the system and protect the control device; The power-off data protection unit provides energy from a backup battery to maintain important data in a specific area when the system loses power. When the power is on, the data is rewritten to the memory or sent to the control center, thereby retaining the last state before the power failure and effectively improving the accuracy of system fault analysis.

8. The system of the intelligent edge computing gateway with asymmetric multi-processor architecture according to claim 1, characterized in that: The encrypted communication includes the following steps: S801, the user logs into the system through the identity key chip, and the hardware true random number generator generates a dynamic key and transmits the information together with the identity key chip information to the connection layer of the cloud control center; S802, the cloud control center connection layer dynamically generates a symmetric exchange key through asymmetric key encryption functions such as ECC and RSA of the data encryption module, and transmits the information to the cloud control center; S803, the cloud control center connection layer receives feedback from the cloud control center and performs AES encryption or decryption processing based on the symmetrical exchange of keys according to the content of the information; S804 The information processed by AES encryption or decryption enters the logic system and enters the shadow storage through the firmware to generate plaintext data.

9. The system of the intelligent edge computing gateway with asymmetric multi-processor architecture according to claim 3, characterized in that: The logic database is used to store pre-compiled logic data in the device and logic data function modules downloaded remotely; The preset function module is used to include multiple integrated functions.

10. An intelligent edge computing gateway with an asymmetric multi-processor architecture, comprising an integrated circuit PCB board, and the following structures integrated on the integrated circuit PCB board: a main memory, a link arbitrator, a dual-port shared memory, an embedded processor and a microcontroller: The main memory includes interpretation support for multiple scripting language programs, implements variable embedded user code, and is used to store data and programs and execute programs; The link arbiter is used to coordinate access when the embedded processor and the microcontroller simultaneously access the same address of the dual-port shared memory, ensure the correct access result, respond to interrupt requests through a priority mechanism, and select the working order of the embedded processor and the microcontroller according to the preset priority; Dual-port shared memory is used for data exchange between embedded processor and microcontroller; The embedded processor is used to parse the script language and transmit it to the microcontroller to complete the cloud control center connection layer. It has an encryption chip, hardware true random number generator, shadow storage, encryption authentication certificate, and firmware disaster recovery function to achieve cloud connection and encrypted communication. The microcontroller is used to receive and execute the script language parsed by the embedded processor, and can cooperate with the embedded processor to complete the variable logic matrix control, connect with the physical device, realize the execution of the physical execution device, realize the offline execution management of the physical device, and use the battery for the internal device; In which, the main memory stores a computer program that can be executed by at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute a system of an intelligent edge computing gateway with an asymmetric multi-processor architecture as described in any one of claims 1 to 9.

Images