An intelligent unmanned AI autonomous control system for work parameters of an engineering machine
By deploying an AI autonomous parameter adjustment terminal and a cloud platform collaborative architecture on construction machinery, and using neural network models for real-time data processing and model updates, the problem of unmanned autonomous control of construction machinery has been solved, realizing unmanned operation in all scenarios and working conditions, and improving work efficiency and equipment reliability.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HUBEI QINGYAN YUNSI INTELLIGENT TECHNOLOGY CO LTD
- Filing Date
- 2026-03-03
- Publication Date
- 2026-06-12
AI Technical Summary
The existing operating parameters of construction machinery cannot be effectively controlled autonomously without human intervention, and they are difficult to adapt to complex working scenarios, resulting in low operating efficiency, high energy consumption, and easy wear and tear on equipment and components.
A hierarchical collaborative architecture based on edge computing AI autonomous parameter tuning terminals and cloud platforms is constructed. Real-time data collection and processing are carried out through AI autonomous parameter tuning terminals deployed on the side of engineering machinery, and the data is uploaded to the cloud platform for training and updating. Real-time control and model updates are carried out using models such as multilayer perceptron neural networks, recurrent neural networks and long short-term memory neural networks.
It enables unmanned operation of construction machinery in all scenarios and working conditions, improving work efficiency, reducing labor costs and energy consumption, and reducing abnormal damage to equipment and components.
Smart Images
Figure CN122194876A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent control technology for construction machinery, and in particular to an intelligent unmanned AI autonomous control system for the operating parameters of construction machinery. Background Technology
[0002] With the decline in the working-age population and the increasing demands for refined management in engineering operations, construction machinery is gradually transforming towards intelligent and unmanned operation. However, due to the complex operating environment and harsh working conditions of construction machinery, and the fact that the intelligent and unmanned process of construction machinery is still in its initial stage, the degree of unmanned operation and intelligence of construction machinery is generally not high. It can only be applied to certain specific simple operation links, and human intervention is still required from time to time. This is accompanied by a series of problems such as low operation efficiency, high energy consumption, and easy wear and tear of equipment and components. Summary of the Invention
[0003] In view of this, it is necessary to provide an intelligent unmanned AI autonomous control system for the operating parameters of construction machinery, so as to effectively solve the technical problems that existing construction machinery operating parameters cannot be effectively controlled autonomously without human intervention and cannot adapt to complex working scenarios.
[0004] This invention provides an intelligent unmanned AI autonomous control system for the operating parameters of construction machinery, including an AI autonomous parameter adjustment terminal deployed on the construction machinery and a cloud platform;
[0005] The AI autonomous parameter tuning terminal is used to collect and analyze data from engineering machinery to obtain time series data. The time series data is stored in its original and complete form and is refreshed periodically. The AI autonomous parameter tuning terminal is equipped with multiple neural network models for time series signal processing. The parsed time series data is input into the neural network model and predictive control results are generated to realize real-time intelligent control of operation parameters.
[0006] The AI autonomous parameter tuning terminal uploads the parsed time series data to the cloud platform. The cloud platform receives, cleans, and labels the received time series data to form a dataset. The cloud platform also deploys multiple neural network models for time series data processing. The dataset is used to train each neural network model, and the trained neural network model is sent back to the AI autonomous parameter tuning terminal to update its neural network model.
[0007] Preferably, the AI autonomous parameter tuning terminal is used to collect and analyze data from the construction machinery to obtain time series data, specifically:
[0008] The AI autonomous parameter tuning terminal accesses multiple bus data and analog data in parallel through a multi-protocol heterogeneous bus processing layer. By adjusting the baud rate to adapt to the bus data of different models, it obtains the original data queue. The original data queue is decoupled from the parsing queue, and the CPU priority of multiple parsing threads is dynamically adjusted according to the load to achieve time synchronization of multi-dimensional data and obtain time series data.
[0009] Preferably, the time series data is stored in its original, complete form and periodically refreshed, specifically as follows:
[0010] The AI autonomous parameter tuning terminal has an embedded solid-state drive, which stores the time-series data in its original and complete form; a data retention mechanism is set up to continuously save data within the most recent period; after confirming that the data has been uploaded to the cloud platform, local data is selectively deleted according to the storage strategy.
[0011] Preferably, the AI autonomous parameter tuning terminal is deployed with multiple neural network models for time-series signal processing, specifically:
[0012] The AI autonomous parameter tuning terminal is equipped with a multilayer perceptron neural network (MLP), a recurrent neural network (RNN), and a long short-term memory neural network (LSTM).
[0013] Preferably, the AI autonomous parameter tuning terminal uploads the parsed time series data to the cloud platform, specifically as follows:
[0014] The AI autonomous parameter tuning terminal adopts a block transmission mechanism, which divides data files larger than a set threshold into fixed-size data packets, compresses the transmitted data, and then uploads the data using the HTTP transmission protocol.
[0015] Preferably, the cloud platform receives and stores the received time-series data to form a dataset, specifically as follows:
[0016] The cloud platform receives data uploaded by the AI autonomous parameter tuning terminal through a dedicated HTTP protocol, and monitors the data transmission status in real time during the reception process to form a data reception log;
[0017] The received data is verified, and once the verification is successful, the data is aggregated into a dataset with a unified format and stored in a streaming database.
[0018] Preferably, the cloud platform also deploys multiple neural network models for time-series signal processing, specifically:
[0019] The neural network models deployed on the cloud platform correspond one-to-one with the neural network models deployed on the AI autonomous parameter tuning terminal.
[0020] Preferably, the neural network models are trained using the dataset, and the trained neural network models are then sent to the AI autonomous parameter tuning terminal to update their neural network models. Specifically:
[0021] The dataset is cleaned and labeled to obtain a training set; the neural network model deployed on the cloud platform is trained using the training set; the phased trained neural network model is then distributed to the AI autonomous parameter tuning terminal to update the neural network model on the AI autonomous parameter tuning terminal.
[0022] Preferably, the AI autonomous parameter tuning terminal continuously collects, parses, and uploads data, continuously trains the neural network model deployed on the cloud platform, sets the update frequency according to actual needs, and periodically updates the neural network model deployed on the AI autonomous parameter tuning terminal according to the update frequency.
[0023] Compared with existing technologies, the advantages of this invention are as follows: Based on an edge computing-based AI autonomous parameter tuning terminal and a cloud platform server, this invention constructs a layered collaborative architecture that enables real-time control of the edge AI model and routine updates of the cloud platform AI model, forming an integrated intelligent processing system. The AI autonomous parameter tuning terminal has multi-bus data acquisition and analysis capabilities and a built-in AI neural network model. By inputting the parsed data into the neural network model, real-time control parameters can be obtained. The cloud platform also deploys a neural network model, receives the data parsed and uploaded by the AI autonomous parameter tuning terminal, continuously learns and trains the neural network model, enabling the model to evolve continuously, and routinely updates the model and distributes it to the AI autonomous parameter tuning terminal. Attached Figure Description
[0024] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:
[0025] Figure 1 is an architecture diagram of an embodiment of an intelligent unmanned AI autonomous control system for engineering machinery operation parameters provided by the present invention. Detailed Implementation
[0026] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.
[0027] Example 1
[0028] To overcome the limitations of existing construction machinery operating parameters, such as their inability to be effectively controlled autonomously without human intervention and their inability to adapt to complex working scenarios, this embodiment proposes an intelligent unmanned AI autonomous control system for construction machinery operating parameters. Please refer to [link / reference needed]. Figure 1 The intelligent unmanned AI autonomous control system for construction machinery operation parameters in this embodiment includes an AI autonomous parameter adjustment terminal deployed on the construction machinery side and a cloud platform;
[0029] The AI autonomous parameter tuning terminal is used to collect and analyze data from engineering machinery to obtain time series data. The time series data is stored in its original and complete form and is refreshed periodically. The AI autonomous parameter tuning terminal is equipped with multiple neural network models for time series signal processing. The parsed time series data is input into the neural network model and predictive control results are generated to realize real-time intelligent control of operation parameters.
[0030] The AI autonomous parameter tuning terminal uploads the parsed time series data to the cloud platform. The cloud platform receives and stores the time series data to form a dataset. The dataset is then cleaned and labeled to obtain a data training set. The cloud platform also deploys multiple neural network models for time series data processing. Each neural network model is trained using the data training set, and the trained neural network model is then sent back to the AI autonomous parameter tuning terminal to update its neural network model.
[0031] In this embodiment, an unmanned real-time AI autonomous parameter tuning terminal is deployed on the construction machinery, and an AI algorithm model, namely a neural network model, is deployed on a cloud platform. Based on a layered collaborative architecture of "real-time control of the edge AI model + routine updates of the cloud platform AI model," operational parameters are continuously collected and uploaded during the operation of the construction machinery. The AI model deployed on the cloud platform is then routinely trained and iteratively optimized, gradually expanding its applicability to different operational scenarios and complex working conditions. The AI model in the edge AI autonomous parameter tuning terminal is remotely upgraded and updated according to actual needs. This achieves unmanned autonomous control of the construction machinery's operational parameters while simultaneously enabling the continuous evolution and improvement of the unmanned operation functions of the construction machinery. Ultimately, this achieves unmanned operation of the construction machinery across all scenarios and working conditions, effectively improving operational efficiency, reducing labor and energy costs, and minimizing abnormal damage to equipment and components.
[0032] This embodiment mainly includes two parts: an edge AI autonomous parameter tuning terminal deployed on construction machinery and a cloud platform. The following will describe these two parts in detail.
[0033] The AI-driven autonomous parameter tuning terminal specifically implements the following functions:
[0034] 1. Data Acquisition and Analysis
[0035] The AI autonomous parameter tuning terminal accesses multiple bus data streams and analog data in parallel through a multi-protocol heterogeneous bus processing layer, supporting dynamic baud rate adjustment to ensure compatibility with different bus models. It decouples the raw data queue from the parsing queue and dynamically adjusts the CPU priority of multiple parsing threads based on the load, achieving microsecond-level time synchronization. This ensures data accuracy in the time dimension, providing reliable time-series data for subsequent AI model calculations, i.e., neural network models.
[0036] 2. Data storage and scrolling refresh
[0037] The AI-driven autonomous parameter tuning terminal has an embedded solid-state drive, storing all collected data in its original, complete form to ensure unfiltered data and no information loss. Simultaneously, a local data retention mechanism is set up, with rolling storage retaining the most recent data to prevent insufficient storage space from preventing the storage of new data. Once the data is uploaded to the cloud platform and successfully received, local data can be selectively deleted according to the storage policy.
[0038] 3. AI model autonomous parameter tuning
[0039] Since the feedback and control data of construction machinery are low-dimensional time-series signals, deep learning algorithm models such as multilayer perceptron neural network (MLP), recurrent neural network (RNN), and long short-term memory neural network (LSTM), which are suitable for time-series signal processing, are deployed on the AI autonomous parameter tuning terminal. This module collects feedback signals such as voltage, current, and pressure of construction machinery in real time, and generates predictive control results efficiently and quickly with low computing resources, enabling intelligent regulation of unmanned real-time automated operation parameters of construction machinery.
[0040] 4. Data upload to cloud platform
[0041] The AI-driven autonomous parameter tuning terminal uses the HTTP transmission protocol. To ensure the accuracy and integrity of data transmission, a chunked transmission mechanism is employed, dividing large data files into fixed-size data packets for transmission. Simultaneously, the transmitted data is compressed to reduce data transmission volume, lower network bandwidth usage, and increase transmission speed while maintaining compression efficiency, ensuring rapid data upload to the cloud platform. Data upload supports 4G / 5G and Wi-Fi communication protocols, automatically selecting the optimal transmission method based on network conditions to prioritize data transmission stability and avoid interruptions.
[0042] The cloud platform specifically implements the following functions:
[0043] 1. Data Reception
[0044] The cloud platform receives data uploaded by the AI autonomous parameter tuning terminal through a dedicated HTTP protocol. During the reception process, it monitors the data transmission status in real time, records key information such as data reception time, data identifier, and data size, and forms a data reception log to facilitate subsequent data traceability and problem investigation.
[0045] After the data is received and verified, it is aggregated into a data stream in a unified format and stored in a streaming database. It supports high-concurrency data writing and fast data reading, and can meet the needs of real-time storage of large amounts of data and subsequent fast retrieval. At the same time, it can classify and index the data according to data type, data upload time and other dimensions to improve data query efficiency.
[0046] 2. AI model training and deployment
[0047] Deep learning algorithm models such as multilayer perceptron neural network (MLP), recurrent neural network (RNN), and long short-term memory neural network (LSTM) suitable for time-series signal processing are deployed on the cloud platform. A large amount of data uploaded to the cloud platform by the AI autonomous parameter tuning terminal is cleaned and labeled to form a data training set. The AI model deployed on the cloud platform is trained, and the stage-trained AI model is distributed to the AI autonomous parameter tuning terminal on the engineering machinery side.
[0048] This embodiment is based on a layered collaborative architecture of "real-time control of the end-side model + normalized updates of the cloud platform model". During the operation of the construction machinery, the operation parameters are continuously collected and uploaded, and the AI model deployed on the cloud platform is normally trained and iteratively optimized. It is gradually expanded to be applicable to different operation scenarios and working conditions, and the end-side AI model is remotely upgraded and updated according to actual needs. This achieves the goal of continuously evolving the automated operation function of the construction machinery and making it better with use, and ultimately realizing unmanned operation of construction machinery in all scenarios and working conditions.
[0049] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of the present invention.
Claims
1. An intelligent unmanned AI autonomous control system for the operating parameters of engineering machinery, characterized in that, This includes AI-powered autonomous parameter tuning terminals deployed on construction machinery and cloud platforms; The AI autonomous parameter tuning terminal is used to collect and analyze data from engineering machinery to obtain time series data. The time series data is stored in its original and complete form and is refreshed periodically. The AI autonomous parameter tuning terminal is equipped with multiple neural network models for time series signal processing. The parsed time series data is input into the neural network model and predictive control results are generated to realize real-time intelligent control of operation parameters. The AI autonomous parameter tuning terminal uploads the parsed time series data to the cloud platform. The cloud platform receives, cleans, and labels the received time series data to form a dataset. The cloud platform also deploys multiple neural network models for time series data processing. The dataset is used to train each neural network model, and the trained neural network model is sent back to the AI autonomous parameter tuning terminal to update its neural network model.
2. The intelligent unmanned AI autonomous control system for engineering machinery operation parameters according to claim 1, characterized in that, The AI-driven autonomous parameter tuning terminal is used to collect and analyze data from construction machinery to obtain time-series data, specifically: The AI autonomous parameter tuning terminal accesses multiple bus data and analog data in parallel through a multi-protocol heterogeneous bus processing layer. By adjusting the baud rate to adapt to the bus data of different models, it obtains the original data queue. The original data queue is decoupled from the parsing queue, and the CPU priority of multiple parsing threads is dynamically adjusted according to the load to achieve time synchronization of multi-dimensional data and obtain time series data.
3. The intelligent unmanned AI autonomous control system for engineering machinery operation parameters according to claim 1, characterized in that, The time series data is stored in its original, complete form and periodically refreshed, specifically as follows: The AI autonomous parameter tuning terminal has an embedded solid-state drive, which stores the time-series data in its original and complete form; a data retention mechanism is set up to continuously save data within the most recent period; after confirming that the data has been uploaded to the cloud platform, local data is selectively deleted according to the storage strategy.
4. The intelligent unmanned AI autonomous control system for engineering machinery operation parameters according to claim 1, characterized in that, The AI autonomous parameter tuning terminal is deployed with multiple neural network models for time-series signal processing, specifically: The AI autonomous parameter tuning terminal is equipped with a multilayer perceptron neural network (MLP), a recurrent neural network (RNN), and a long short-term memory neural network (LSTM).
5. The intelligent unmanned AI autonomous control system for engineering machinery operation parameters according to claim 1, characterized in that, The AI-driven autonomous parameter tuning terminal uploads the parsed time-series data to the cloud platform, specifically as follows: The AI autonomous parameter tuning terminal adopts a block transmission mechanism, which divides data files larger than a set threshold into fixed-size data packets, compresses the transmitted data, and then uploads the data using the HTTP transmission protocol.
6. The intelligent unmanned AI autonomous control system for engineering machinery operation parameters according to claim 1, characterized in that, The cloud platform receives and stores the received time-series data to form a dataset, specifically as follows: The cloud platform receives data uploaded by the AI autonomous parameter tuning terminal through a dedicated HTTP protocol, and monitors the data transmission status in real time during the reception process to form a data reception log; The received data is verified, and once the verification is successful, the data is aggregated into a dataset with a unified format and stored in a streaming database.
7. The intelligent unmanned AI autonomous control system for engineering machinery operation parameters according to claim 1, characterized in that, The cloud platform also deploys multiple neural network models for time-series signal processing, specifically: The neural network models deployed on the cloud platform correspond one-to-one with the neural network models deployed on the AI autonomous parameter tuning terminal.
8. The intelligent unmanned AI autonomous control system for engineering machinery operation parameters according to claim 1, characterized in that, The aforementioned dataset is used to train each of the neural network models, and the trained neural network models are then sent to the AI autonomous parameter tuning terminal to update their neural network models. Specifically: The dataset is cleaned and labeled to obtain a training set; the neural network model deployed on the cloud platform is trained using the training set; the phased trained neural network model is then distributed to the AI autonomous parameter tuning terminal to update the neural network model on the AI autonomous parameter tuning terminal.
9. The intelligent unmanned AI autonomous control system for engineering machinery operation parameters according to claim 1, characterized in that, The AI autonomous parameter tuning terminal continuously collects, parses, and uploads data, continuously trains the neural network model deployed on the cloud platform, sets the update frequency according to actual needs, and periodically updates the neural network model deployed on the AI autonomous parameter tuning terminal according to the update frequency.