Multi-sensor fusion optimization system based on micro machine learning

Abstract

The invention discloses a multi-sensor fusion optimization system based on micro machine learning, belongs to the technical field of regional sharing, and aims to solve the technical problems of how to improve the use efficiency of regional resources and reduce the vacancy rate of the resources. Comprising a regional resource management terminal and a regional resource management platform. The regional resource management terminal is used for taking the preprocessed space use state information as input, and predicting and outputting the occupancy rate of corresponding regional space resource use through a trained TinyML model; the regional resource management platform interacts with a regional administrator as a user, a leasing person and a platform administrator, supports the regional administrator and the leasing person to carry out leasing transaction and leasing settlement on a selected idle region, and records a leasing transaction process and a leasing settlement process to form regional leasing information; a platform administrator is supported to carry out identity authentication on a regional administrator and a leasing person who carry out leasing transaction and leasing settlement, and reasonable configuration and efficient utilization of regional resources are realized.

Description

Technical Field

[0001] The present invention relates to the field of regional sharing technology, and in particular to a multi-sensor fusion optimization system based on micro machine learning. Background Art

[0002] With the acceleration of urbanization and economic development, people's demand for efficient use of space resources is growing. Regional resources such as warehouses and private parking spaces are often idle during certain periods of time, resulting in a waste of resources. How to reasonably utilize these regional resources when they are idle and improve their efficiency has become an important issue at present.

[0003] Although there are some solutions for regional resource leasing in the existing technology, these solutions usually have some limitations, such as cumbersome leasing process, lack of effective security mechanism, and inability to dynamically match according to the real-time status of regional resources. These solutions restrict the development of regional resource leasing and cannot meet people's demand for efficient use of spatial resources.

[0004] How to improve the utilization efficiency of regional resources and reduce the idle rate of resources is a technical problem that needs to be solved. Summary of the invention

[0005] The technical task of the present invention is to address the above shortcomings and provide a multi-sensor fusion optimization system based on micro machine learning to solve the technical problem of how to improve the utilization efficiency of regional resources and reduce the idle rate of resources.

[0006] The present invention provides a multi-sensor fusion optimization system based on micro machine learning, including a regional resource management terminal and a regional resource management platform; Each region is configured with a corresponding regional resource management terminal, which stores a model file issued by the regional resource management platform. The regional resource management terminal is used to collect the space usage status information of the corresponding region in real time and pre-process the space usage status information, and is used to deploy the trained TinyML model locally by running the model file, and use the pre-processed space usage status information as input, and predict and output the occupancy rate of the space resource usage of the corresponding region through the trained TinyML model; The regional resource management platform interacts with regional administrators, leasing personnel and platform administrators as users, and is used to support regional administrators in publishing regional information of their idle areas, to support platform administrators in formulating transaction rules and specifying transaction prices for each idle area, to support regional administrators and leasing personnel in conducting leasing transactions and leasing settlements for selected idle areas, and to record the leasing transaction process and the leasing settlement process to form regional leasing information, and to support platform administrators in performing identity authentication on regional administrators and leasing personnel conducting leasing transactions and leasing settlements.

[0007] Preferably, the regional resource management terminal includes a data acquisition unit, a data processing unit and a controller; The data acquisition unit is used to monitor the space usage status of the corresponding area in real time through multi-source sensors, obtain space usage status information in the form of images and data, and send the space usage status information to the controller and the data processing unit; The data processing unit is used to perform data preprocessing on the space usage status information, and to perform data cleaning and formatting on the space usage status information through data preprocessing to obtain the preprocessed space usage status information; The controller stores a model file issued by the regional resource management platform, which is used to deploy the trained TinyML model based on the model file, and take the preprocessed space usage status information as input, predict and output the occupancy rate of the corresponding regional space resource usage through the trained TinyML model.

[0008] Preferably, the controller is also used to adaptively adjust the parameters of the trained TinyML model based on the preprocessed space usage status information and the predicted output space resource usage occupancy rate to obtain an adjusted TinyML model, and is used to use the subsequently received preprocessed space usage status information as input, and predict and output the space resource usage occupancy rate of the corresponding area through the adjusted TinyML model.

[0009] Preferably, the area information includes a unique identification number of the area, area location information, occupancy rate of space resource usage, and idle time, wherein the idle time indicates the time when the area administrator does not use his area.

[0010] Preferably, the regional resource management platform interacts with the user through the regional resource management interface to provide the following services: Supports regional administrators to register and log in, supports regional administrators to upload regional information of their idle areas, and displays the regional information of each idle area through the regional resource management interface; Support leasing personnel to register and log in, and support leasing personnel to browse the regional information of each free area through the regional resource management interface; Support leasing personnel to filter free areas through the regional resource management interface, and support leasing personnel to initiate leasing requests to platform administrators based on the selected free areas, where the leasing requests include the regional information of the designated free areas and the personal verification information of the leasing personnel; Support platform administrators to register and log in, and support platform administrators to formulate transaction rules and price strategies, and set the rental price for each vacant area based on the price strategy; Support platform administrators to review rental personnel based on rental requests and authorize the use of idle areas for rental personnel who have passed the review; For the vacant areas with authorized use, the platform administrator is supported to authenticate the identities of the regional administrators and leasing personnel. After the authentication, the regional administrators and leasing personnel are supported to conduct leasing transactions and leasing settlements, and record the leasing transaction process and leasing settlement process to form regional leasing information. During the leasing transaction, the transaction price is calculated based on the transaction rules formulated by the platform administrator, the regional information specified in the leasing request and the leasing price of the corresponding area. During the leasing settlement, the platform administrator is supported to authenticate the identities of the regional administrators and leasing personnel conducting the leasing transaction, and the leasing personnel are supported to make secure payments based on the transaction price.

[0011] Preferably, the regional resource management platform is used to collect leasing information, including historical regional leasing information, market leasing information, feedback from leasing users on the use of their leased vacant areas, and environmental information of vacant areas, and is used to adjust transaction rules and pricing strategies based on the leasing information, and to optimize the current trained TinyML model, use the optimized TinyML model as the trained TinyML model, and send the model file of the trained TinyML model to each regional resource management terminal.

[0012] Preferably, the regional resource management platform is used to compress and convert the model file of the trained TinyML model into a binary file, and send the binary file to each regional resource management terminal.

[0013] Preferably, the regional resource management platform is a blockchain platform that introduces blockchain, and smart contracts are deployed in the regional resource management platform. The regional leasing information is stored through the blockchain, and the leasing transaction rules, price strategies and leasing prices of each region are stored through smart contracts. Based on the smart contracts, the regional administrators and leasing users who conduct leasing transactions are authenticated, and leasing transactions and leasing settlements are executed.

[0014] The multi-sensor fusion optimization system based on micro machine learning of the present invention has the following advantages: 1. It realizes the rational allocation and efficient utilization of regional resources, which not only improves the economic benefits of regional owners or managers, but also provides convenient and reliable leasing services for leasing users. Through standardized management and intelligent operation, it promotes resource sharing and sustainable development; 2. Record regional leasing information through blockchain, and execute leasing transactions, leasing settlements, and identity authentication of regional administrators and leasing users through smart contracts to ensure data privacy and non-tamperability, thereby enhancing the system's security authentication mechanism; 3. The regional resource management terminal adaptively adjusts the parameters of the trained TinyML model based on the pre-processed space usage status information and the predicted output space resource usage occupancy rate, thus realizing the model's adaptation and improving the prediction accuracy of the interval space resource utilization occupancy rate; 4. The regional resource management platform adjusts the transaction rules and pricing strategies based on the rental information it collects, and optimizes the currently trained TinyML model, improving the accuracy of the model and facilitating better service to rental users. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] The present invention is further described below in conjunction with the accompanying drawings.

[0017] Figure 1 This is a flowchart of the workflow of a multi-sensor fusion optimization system based on micro machine learning in Example 1. DETAILED DESCRIPTION

[0018] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it. However, the embodiments are not intended to limit the present invention. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments may be combined with each other.

[0019] The embodiments of the present invention provide a multi-sensor fusion optimization system and method based on micro machine learning, which are used to solve the technical problem of how to improve the utilization efficiency of regional resources and reduce the idle rate of resources.

[0020] Embodiment 1: The present invention provides a multi-sensor fusion optimization system based on micro machine learning, including a regional resource management terminal and a regional resource management platform.

[0021] Each area is configured with a corresponding regional resource management terminal, which stores the model file issued by the regional resource management platform. The regional resource management terminal is used to collect the space usage status information of the corresponding area in real time and pre-process the space usage status information. It is used to deploy the trained TinyML model locally by running the model file, and use the pre-processed space usage status information as input, and predict and output the occupancy rate of the space resources in the corresponding area through the trained TinyML model.

[0022] As a specific implementation of the regional resource management terminal, it includes a data acquisition unit, a data processing unit and a controller.

[0023] The data acquisition unit is used to monitor the space usage status of the corresponding area in real time through multi-source sensors, obtain space usage status information in the form of images and data, and send the space usage status information to the controller and the data processing unit.

[0024] The data processing unit is used to perform data preprocessing on the space usage status information, and through data preprocessing, the space usage status information is cleaned and formatted to obtain the preprocessed space usage status information.

[0025] The controller stores the model file issued by the regional resource management platform, which is used to deploy the trained TinyML model based on the model file, and take the preprocessed space usage status information as input, predict and output the occupancy rate of the corresponding regional space resources through the trained TinyML model.

[0026] Taking into account that each regional resource management end is oriented to different regions, in order to adapt to the accurate prediction of the space resource utilization in each region, the controller is also used to adaptively adjust the parameters of the trained TinyML model based on the preprocessed space usage status information and the predicted output space resource usage occupancy rate, and obtain the adjusted TinyML model, and is used to predict and output the space resource usage occupancy rate of the corresponding region through the adjusted TinyML model with the subsequently received preprocessed space usage status information as input. The controller adjusts the parameters of the current trained TinyML model based on the subsequently collected preprocessed space usage status information and the predicted space resource usage occupancy rate, realizing the adaptive adjustment of the model and improving the accuracy of the prediction.

[0027] The regional resource management platform interacts with regional administrators, leasing personnel and platform administrators as users, and is used to support regional administrators in publishing regional information of their idle areas, to support platform administrators in formulating transaction rules and specifying transaction prices for each idle area, to support regional administrators and leasing personnel in conducting leasing transactions and leasing settlements for selected idle areas, and to record the leasing transaction process and leasing settlement process to form regional leasing information, and to support platform administrators in performing identity authentication on regional administrators and leasing personnel conducting leasing transactions and leasing settlements.

[0028] The area information includes the area unique identification number, area location information, space resource usage occupancy rate and idle time, wherein the idle time indicates the time when the area administrator does not use his area.

[0029] The controller's on-chip storage resources are limited, so the model files need to be compressed, and the model files stored on the controller must be binary files. Therefore, the regional resource management platform is used to compress and convert the model files of the trained TinyML model into binary files, and send the binary files to each regional resource management terminal.

[0030] In this embodiment, the regional administrator can be the regional user or a third party entrusted by the regional user to manage the area. The regional administrator can customize the usage time of his area, and the remaining time is the idle time of the area. The area can be rented out during the idle time to achieve regional resource sharing.

[0031] As a specific implementation of the regional resource management platform, the regional resource management platform is a blockchain platform that introduces blockchain. Smart contracts are deployed in the regional resource management platform. The regional leasing information is stored through blockchain, and the leasing transaction rules, price strategies and leasing prices of each region are stored through smart contracts. Based on smart contracts, regional administrators and leasing users who conduct leasing transactions are authenticated, and leasing transactions and leasing settlements are executed. The platform interacts with users through the regional resource management interface to provide the following services: (1) Supports regional administrators to register and log in, supports regional administrators to upload regional information of their idle areas, and displays the regional information of each idle area through the regional resource management interface; (2) Support leasing personnel to register and log in, and support leasing personnel to browse the regional information of each idle area through the regional resource management interface; (3) Support leasing personnel to screen idle areas through the regional resource management interface, and support leasing personnel to initiate leasing requests to platform administrators based on the selected idle areas, where the leasing requests include regional information of the designated idle areas and personal verification information of the leasing personnel; (4) Support platform administrators to register and log in, and support platform administrators to formulate transaction rules and pricing strategies, and set the rental price for each vacant area based on the pricing strategy; (5) Support platform administrators to review rental personnel based on rental requests and authorize the use of idle areas for rental personnel who have passed the review; (6) For vacant areas that have been authorized for use, the platform administrator is supported to authenticate the identities of regional administrators and leasing personnel. After authentication, the regional administrators and leasing personnel are supported to conduct leasing transactions and leasing settlements, and the leasing transaction process and leasing settlement process are recorded to form regional leasing information. During leasing transactions, the transaction price is calculated based on the transaction rules formulated by the platform administrator, the regional information specified in the leasing request, and the leasing price of the corresponding area. During leasing settlement, the platform administrator is supported to authenticate the identities of regional administrators and leasing personnel conducting leasing transactions, and the leasing personnel are supported to make secure payments based on the transaction price.

[0032] For the system disclosed in this embodiment, regional information collection can be realized through the data collection unit in the regional resource management terminal, mainly collecting data information of the current site area, which can be collected by a camera to collect current image information, or by a regional detection sensor to detect data information of the current site area, and transmit the collected data information of the current site area to the data processing unit. The data processing unit pre-processes the collected space usage status information, takes the pre-processed space usage status information as input, and predicts and outputs the occupancy rate of the corresponding regional space resource usage through the trained TinyML model deployed on the controller, thereby assisting the regional administrator to determine the usage of the corresponding area and whether it is an idle area available for rent.

[0033] In this embodiment, on the basis of the existing micro machine learning model optimization, a self-optimization mechanism is introduced, which can dynamically adjust the model parameters according to real-time data, which will improve the flexibility and adaptability of the model, especially when the resource usage pattern changes, the model can respond quickly and adjust the strategy. By monitoring the changes in resource status, the micro machine learning model can be updated regularly, and the most appropriate parameters and algorithms can be automatically selected to optimize resource allocation. This method not only improves the efficiency of resource use, but also enhances the system's adaptability to different scenarios. A large amount of data that meets the scenario is selected, and a data set that meets the application scenario is produced. The collected large amount of data set information is used for model training, and the trained model is optimized to the best state. The on-chip storage resources of the microcontroller are limited, and the model needs to be compressed. The model file deployed on the microcontroller must be a binary file, so the compressed model file needs to be processed as a binary file. After completion, the binary file is deployed to the microcontroller, and the microcontroller runs the model file.

[0034] For the system of this embodiment, the following services can be provided through the regional resource management platform: (1) Automatically publish idle information. Regional managers enter their own usage information into the management platform, and the management platform automatically publishes the idle time of the area for rental personnel to view the idle time information of the area.

[0035] (2) Security authentication mechanism: Regional managers need to submit all regional-related certification materials to the management platform, while regional rental personnel need to submit their own personal certification materials.

[0036] (3) Dynamic pricing and payment: Dynamic pricing and payment adjustments are made based on market prices and regional characteristics to maximize economic benefits.

[0037] (4) Data management and analysis: Effectively manage and deeply mine leasing data, use advanced data analysis technology and algorithms to analyze, compare and summarize leasing data to reveal the laws and trends behind the data. Through in-depth analysis of the data, we can more accurately understand information on market demand, regional characteristics and user behavior, and provide strong decision-making support for the optimal allocation of regional resources and future development. These data include not only leasing transaction data, but also user feedback, regional environment, market dynamics and other information. Through comprehensive analysis of these data, we can better understand market dynamics and user needs, and provide a scientific basis for the optimal allocation of regional resources. At the same time, through in-depth mining and analysis of data, we can also discover potential market opportunities and commercial value, and provide important decision-making support for the future development of the region.

[0038] The regional resource management platform of this embodiment is an integrated system for managing and monitoring regional leasing business. The platform divides users into three roles: regional administrator, platform administrator and leasing staff.

[0039] The regional administrator is the owner of the region or a third-party company. He is mainly responsible for publishing the idle time periods of the region and earning income. The regional administrator needs to publish the specific location of the region and the purpose of the region in the management platform, such as parking and warehouse storage. Maintaining the regional environment, the regional administrator is responsible for keeping the regional environment clean and safe. They need to regularly inspect and maintain facilities and equipment to ensure the normal operation of the region and provide a good rental environment.

[0040] The platform administrator sets and adjusts the pricing strategy. The platform administrator needs to set reasonable rental prices and related fees based on market demand and regional characteristics. According to market changes and regional demand, the pricing strategy needs to be adjusted in a timely manner to ensure the maximization of economic benefits. However, the price range must meet the price range provided by the platform. Supervision and handling of violations: When violations are found in the rental users, the platform administrator needs to take corresponding measures, such as warnings, fines or termination of contracts. They also need to handle user complaints and disputes in a timely manner to ensure the fairness and legality of the rental process.

[0041] Leasing personnel, that is, individuals or enterprises with leasing needs, mainly find and lease suitable areas through the management platform. Leasing users need to browse and select suitable areas on the platform to view the specific information, purpose, price and off-peak time period of the area. Users can compare and filter on the platform according to their needs and budget, and choose the most suitable area for leasing. User authentication and payment: After selecting the area, the leasing user needs to perform identity authentication to ensure that his identity is authentic and reliable. After passing the authentication, the leasing user needs to pay the corresponding fee according to the set price. The payment method should be safe and convenient, and the platform should provide a variety of payment options to meet the needs of different users. Comply with leasing regulations: Leasing users must comply with the leasing regulations of the management platform and regional administrators, and must not use the leasing area for illegal purposes. Leasing users must also use and maintain the leasing area reasonably to ensure its normal operation and extend its service life. Timely feedback and evaluation: After using the area, leasing users should promptly feedback their experience and evaluation to the platform. This helps regional administrators improve service quality and increase user satisfaction.

[0042] In this embodiment, the platform adds smart contracts and blockchain technology to ensure the transparency and immutability of the leasing process. Each resource leasing and payment process can be recorded through the blockchain to ensure the security and credibility of the transaction between the two parties. After the leasing transaction is completed, the smart contract automatically settles and verifies the identities and transactions of both parties, ensuring the privacy and immutability of the data, thereby enhancing the security authentication mechanism of the system.

[0043] Through the unified management and operation of the management platform, the rational allocation and efficient utilization of regional resources are achieved. It not only improves the economic benefits of regional owners or managers, but also provides convenient and reliable leasing services for leasing users. In addition, the platform promotes resource sharing and sustainable development through standardized management and intelligent operation.

[0044] The system disclosed in this embodiment can realize the leasing of idle areas, and the specific operations are as follows: Step S100, for each area, the regional resource management terminal configured in the area performs the following: real-time collection of space usage status information of the corresponding area and preprocessing of the space usage status information, local deployment of the trained TinyML model by running the model file, and using the preprocessed space usage status information as input, predicting and outputting the occupancy rate of space resource usage in the corresponding area through the trained TinyML model, and judging whether the area is an idle area based on the occupancy rate; Step S200: The regional administrator publishes the block information of its idle area to the regional resource management platform; Step S300: The platform administrator formulates transaction rules and specifies a transaction price for each free area; Step S400: The regional administrator and the leasing staff conduct leasing transactions and leasing settlements for the selected idle areas, and record the leasing transaction process and the leasing settlement process to form regional leasing information. The platform administrator performs identity authentication on the regional administrator and the leasing staff who conduct the leasing transactions and leasing settlements.

[0045] The system of this embodiment can be applied to areas such as warehouses and parking spaces. Here is a specific example: a parking space owner may use the parking space from 6 pm to 7 am, and the rest of the time the parking space is idle. If the idle time is not adjusted, it is easy to cause a waste of this period of resources. The parking space owner can rent out the idle parking space to others without affecting his own use, thereby realizing effective utilization of resources.

[0046] Parking space owners publish their parking space idle time information through the management platform. This information includes the specific location, size, safety status, rent, etc. The published information will be automatically pushed to the rental platform for potential renters to view and select.

[0047] Renters browse and select parking spaces that suit them through the management platform. They can filter and compare according to their needs, such as parking time, price budget, etc.

[0048] Once the renter has selected a suitable parking space and completed the payment, the management platform will automatically send an electronic key to the renter's mobile phone. This electronic key can be used to unlock the parking space within the specified time.

[0049] When the lessee needs to end the lease, they only need to operate on the management platform, and the platform will automatically notify the parking space owner and complete the rent settlement.

[0050] In order to ensure the safety of both parties, the management platform has introduced an intelligent security authentication mechanism. All users operating on the platform need to pass real-name authentication to ensure that their identities are authentic and reliable. In addition, the platform will monitor the entire leasing process to ensure that the transaction is fair and legal.

[0051] The management platform also regularly collects and analyzes rental data to provide more personalized services to parking space owners and renters. For example, based on rental history and demand forecasts, the management platform can provide more reasonable pricing strategies for receipt management users to obtain higher returns.

[0052] The above is a detailed introduction to a multi-sensor fusion optimization system based on micro machine learning provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technicians in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A multi-sensor fusion optimization system based on micro machine learning, characterized in that: Including regional resource management terminal and regional resource management platform; Each region is configured with a corresponding regional resource management terminal, which stores a model file issued by the regional resource management platform. The regional resource management terminal is used to collect the space usage status information of the corresponding region in real time and pre-process the space usage status information, and is used to deploy the trained TinyML model locally by running the model file, and use the pre-processed space usage status information as input, and predict and output the occupancy rate of the space resource usage of the corresponding region through the trained TinyML model; The regional resource management platform interacts with regional administrators, leasing personnel and platform administrators as users, and is used to support regional administrators in publishing regional information of their idle areas, to support platform administrators in formulating transaction rules and specifying transaction prices for each idle area, to support regional administrators and leasing personnel in conducting leasing transactions and leasing settlements for selected idle areas, and to record the leasing transaction process and the leasing settlement process to form regional leasing information, and to support platform administrators in performing identity authentication on regional administrators and leasing personnel conducting leasing transactions and leasing settlements.

2. The multi-sensor fusion optimization system based on micro machine learning according to claim 1 is characterized in that: The regional resource management terminal includes a data acquisition unit, a data processing unit and a controller; The data acquisition unit is used to monitor the space usage status of the corresponding area in real time through multi-source sensors, obtain space usage status information in the form of images and data, and send the space usage status information to the controller and the data processing unit; The data processing unit is used to perform data preprocessing on the space usage status information, and to perform data cleaning and formatting on the space usage status information through data preprocessing to obtain the preprocessed space usage status information; The controller stores a model file issued by the regional resource management platform, which is used to deploy the trained TinyML model based on the model file, and take the preprocessed space usage status information as input, predict and output the occupancy rate of the corresponding regional space resource usage through the trained TinyML model.

3. The multi-sensor fusion optimization system based on micro machine learning according to claim 1 is characterized in that: The controller is also used to adaptively adjust the parameters of the trained TinyML model based on the preprocessed space usage status information and the predicted output space resource usage occupancy rate to obtain an adjusted TinyML model, and is used to use the subsequently received preprocessed space usage status information as input, and predict and output the space resource usage occupancy rate of the corresponding area through the adjusted TinyML model.

4. The multi-sensor fusion optimization system based on micro machine learning according to claim 1 is characterized in that: The area information includes the area unique identification number, area location information, occupancy rate of space resource usage, and idle time, where the idle time indicates the time when the area administrator does not use his area.

5. The multi-sensor fusion optimization system based on micro machine learning according to claim 1 is characterized in that: The regional resource management platform interacts with users through the regional resource management interface to provide the following services: Supports regional administrators to register and log in, supports regional administrators to upload regional information of their idle areas, and displays the regional information of each idle area through the regional resource management interface; Support leasing personnel to register and log in, and support leasing personnel to browse the regional information of each free area through the regional resource management interface; Support leasing personnel to filter free areas through the regional resource management interface, and support leasing personnel to initiate leasing requests to platform administrators based on the selected free areas, where the leasing requests include the regional information of the designated free areas and the personal verification information of the leasing personnel; Support platform administrators to register and log in, and support platform administrators to formulate transaction rules and price strategies, and set the rental price for each vacant area based on the price strategy; Support platform administrators to review rental personnel based on rental requests and authorize the use of idle areas for rental personnel who have passed the review; For the vacant areas with authorized use, the platform administrator is supported to authenticate the identities of the regional administrators and leasing personnel. After the authentication, the regional administrators and leasing personnel are supported to conduct leasing transactions and leasing settlements, and record the leasing transaction process and leasing settlement process to form regional leasing information. During the leasing transaction, the transaction price is calculated based on the transaction rules formulated by the platform administrator, the regional information specified in the leasing request and the leasing price of the corresponding area. During the leasing settlement, the platform administrator is supported to authenticate the identities of the regional administrators and leasing personnel conducting the leasing transaction, and the leasing personnel are supported to make secure payments based on the transaction price.

6. The multi-sensor fusion optimization system based on micro machine learning according to claim 1 is characterized in that: The regional resource management platform is used to collect leasing information, including historical regional leasing information, market leasing information, feedback from leasing users on the use of their leased vacant areas, and environmental information of vacant areas, and is used to adjust transaction rules and pricing strategies based on the leasing information, and to optimize the current trained TinyML model, use the optimized TinyML model as the trained TinyML model, and send the model file of the trained TinyML model to each regional resource management terminal.

7. The multi-sensor fusion optimization system based on micro machine learning according to claim 1 is characterized in that: The regional resource management platform is used to compress and convert the model files of the trained TinyML model into binary files, and send the binary files to each regional resource management terminal.

8. The multi-sensor fusion optimization system based on micro machine learning according to claim 1 is characterized in that: The regional resource management platform is a blockchain platform that introduces blockchain. Smart contracts are deployed in the regional resource management platform. Regional leasing information is stored through blockchain, and leasing transaction rules, price strategies and leasing prices for each region are stored through smart contracts. Based on smart contracts, regional administrators and leasing users who conduct leasing transactions are authenticated, and leasing transactions and leasing settlements are executed.

Images