TRICONEX 3503EN、3503E、3502E、3501E、3481、3451、3351、3401、3301、3201、3101、3009、3008、3006、3005、3002、2402

TRICONEX 3503EN、3503E、3502E、3501E、3481、3451、3351、3401、3301、3201、3101、3009、3008、3006、3005、3002、2402TRICONEX 3503EN、3503E、3502E、3501E、3481、3451、3351、3401、3301、3201、3101、3009、3008、3006、3005、3002、2402TRICONEX 3503EN、3503E、3502E、3501E、3481、3451、3351、3401、3301

3201、3101、3009、3008、3006、3005、3002、2402TRICONEX 3503EN、3503E、3502E、3501E、3481、3451、3351、3401、3301、3201、3101、3009、3008、3006、3005、3002、2402TRICONEX 3503EN、3503E、3502E、3501E、 The motion trajectory planning of cars 3481, 3451, and 3351, the control algorithm of the steering wheel motor, and the coordinate system transformation between the motor and the car are all programmed and implemented in Automation Studio.

Scene Viewer is a 3D graphics display software that allows you to view the dynamic running effects of the final car in Scene Viewer. When the control strategy is not properly designed, an alarm can also be seen in the Scene Viewer for AGV hitting the boundary when turning. The Scene Viewer software here is an open-source software.

MapleSim software is built on top of the famous mathematical software Maple. Maple software has powerful numerical and symbolic computing capabilities, especially Maple’s symbolic computing, which is the derivation of mathematical formulas. It is second to none in the world, such as calculus, solving general and special solutions of differential equations, etc. Maple and MapleSim are bridge tools between classroom teaching and practice.

The control object in the competition is an AGV car, and we cannot use an actual car in the design. We can only use a simulation model instead. The AGV car model includes a mechanism, a motor, and wheels, which are driven by the motor to move on a designated track. This model is connected through the interface between MapleSim and Automation Studio, achieving integrated closed-loop testing of control algorithms and controlled object models.

We have provided an AGV simulation model for the competition problem, and the software’s built-in help system lists the underlying mathematical equations of different modeling components. Through these, everyone can understand the model principles, and then practice, modify and improve the model. I believe everyone can achieve success! Firstly, the AGV model used in this competition adopts the simplest three wheel structure, with one being a single steering wheel and the other two being passive wheels.

In the two motors of a single steering wheel, one is responsible for steering and the other is responsible for driving. In addition, we have already built the mechanical physics model of the AGV in MapleSim and converted it into FMU, which has been imported into Automation Studio. At the same time, we have provided a simple control model demo in Automation Studio to help everyone run the mechanical and control models together.

In the final open question, it involves modifying the AGV physical model. At this point, everyone needs to open MapleSim, modify the friction parameters in the corresponding module, and then import the new model into Automation Studio to verify the adaptability of the control algorithm to the new model. The open question will have additional points. In addition, for the three loop control of the motor, there are two methods that can be used to solve it. One is to build control models for the position loop, speed loop, and current loop in Automation Studio, and the other is to directly call on the servo drive of B&R, so that the control model does not need to be built by oneself, and only the parameters need to be adjusted and configured appropriately. We recommend the first method here, and if students can implement both methods, there will be additional points.

In addition, for the three loop control of the motor, there are two methods that can be used to solve it. One is to build control models for the position loop, speed loop, and current loop in Automation Studio, and the other is to directly call on the servo drive of B&R, so that the control model does not need to be built by oneself, and only the parameters need to be adjusted and configured appropriately. We recommend the first method here, and if students can implement both methods, there will be additional points.

Step 1- Familiarize yourself with the modeling and simulation software MapleSim: refer to the manual and video materials, practice hands-on, and understand the modeling methods and principles. The focus of this AGV car model is on the dynamic simulation part. If the principles of the model, especially the wheel model, can be explained and explained, I believe it will give extra points to the answer

Step 2- Adjust Model Parameters: According to the requirements of the competition, adjust the friction parameters of the contact elements in the wheels. Through parameter scanning and optimization, simulate the kinematic and dynamic characteristics of the AGV car under different parameter conditions.

Step 3- Output the simulation model to the control software Automation Studio: Use the Automation Studio FMU interface toolbox in MapleSim to output the modified model as an FMU file. Debug the AGV trolley’s motion attitude in Automation Studio and verify the control strategy

Secondly, the competition requires participants to have a solid theoretical foundation in mechatronics, understand the servo motor control model in motion control, and know how to adjust the three loop control parameters of the servo drive to improve the system’s fast response ability, stability, and control accuracy. We also need to study how to coordinate transform the motor position coordinate system with the AGV position coordinate system.

Thirdly, two of the tasks in this competition involve optimization theory, one is to minimize time and the other is to minimize energy consumption, both of which have corresponding scenarios in real life.

Recently, ABB provided leading electrical solutions and i-bus intelligent building control systems for Dongguan International Trade Center. At the equipment level, create a safe, reliable, and stable power distribution system through low-voltage equipment; At the control layer, an i-bus system is provided to achieve scene based lighting control in public areas, creating smarter and more comfortable living spaces while meeting green and low-carbon environmental requirements and reducing building carbon emissions. In the project, involving over 1000 circuits, it is expected to save more than 30% energy compared to traditional lighting solutions. The International Trade Center project is positioned as a first tier urban complex with a high electricity load and a variety of electrical equipment. The super high-rise buildings have a huge volume and complex functions, and the selection of power distribution systems needs to meet stricter standards. According to customer needs, ABB has sorted out a low-voltage component product portfolio consisting of air circuit breakers, molded case circuit breakers, miniature circuit breakers, dual power transfer switches, isolating switch fuses, etc. Even under heavy loads, it can ensure the safe and stable operation of the distribution system in the International Trade Center.

oplus_10485760

In addition, the client proposed that the building intelligent control system not only needs to have basic functions such as lighting control, but also needs to have stability, energy efficiency, and scalability to ensure keeping up with the times. ABB i-bus ® The intelligent building control system sets scenarios for different areas and time periods through real-time monitoring, scene control, timed control, and other functions to meet customers’ requirements for energy conservation and consumption reduction. At the same time, ABB i-bus ® The intelligent building control system is integrated with the building automation system through fiber optic transmission scheme, and remotely controls the lighting in different areas in the monitoring center, meeting customers’ requirements for unified control of various scenarios in public areas, and enabling the comprehensive intelligent upgrade of the International Trade Center tower. Retrieve and output instructions (LD/LDI/LDP/LDF/OUT)

In March 2016, the game between AlphaGo and professional Go players sparked high attention to artificial intelligence. Did computers rely on human like intelligence to defeat the top players in a widely recognized and highly complex computational and intellectual task? From the perspective of system structure, AlphaGo combines deep neural network training with Monte Carlo simulation [1]. In a broad sense, deep neural networks are a computational form similar to the brain, while Monte Carlo methods leverage the advantage of machine processing speed to simulate a vast number of possibilities for further judgment, which now appears to be not the mechanism of brain operation. So AlphaGo can be said to have achieved success by combining the computing and intelligence of both brain like and non brain like systems, perfectly leveraging their respective strengths. What other aspects of brain like computing and intelligence are currently being studied besides the deep neural networks used by AlphaGo? What kind of breakthrough may it bring in the near future?

1. Rich interfaces, supporting device access such as Ethernet, serial port, CAN port, IO port, as well as Ethernet, 2G/3G/4G full network network access;

2. Embedded with hundreds of industrial protocols, supporting over 99% of PLCs and the vast majority of industrial equipment access;

3. 8GB local storage+SD card support, supporting local data caching and offline applications;

4. Three in one serial port, supporting RS485/RS232/RS422 electrical interfaces;

5. Support edge computing, realize data optimization, real-time response, agile connection, model analysis and other services at the edge node of the Internet of Things, and effectively share cloud computing resources to support simultaneous access of multiple devices;

6. Supports DC9~36V wide voltage input, suitable for various complex industrial sites;

7. Support LED lamp customization, and users can define LED lamps as required (such as device status, edge computing results, etc.);

8. No need for a client, supports on-demand remote upload and download, effectively saving network traffic;

9. Support gateway health self diagnosis and quick detection of gateway faults;

10. Supports multiple standard VPNs (PPTP/L2TP/IPSec/OpenVPN);

11. Support network primary backup mode, intelligently switch network access methods based on network conditions (support intelligent network diagnosis);

12. Powerful cloud software center support, which can install corresponding firmware, applications, etc. according to actual application scenarios;

13. Supports multiple remote control modes (no password/with password/disabled), and also has physical remote control switch and one key switch remote control function;

14. Support multi link well data collection;

15. Support 4G data detail analysis and data control;

16. Support gateway remote management; Support network self recovery;

17. Support mixed positioning mode of base station and GPS, as well as local WEB GPS location presentation;

18. Support local WEB endpoint table configuration, support local configuration design and presentation;

19. Industrial edge computing gateway, with a maximum of 5000 data acquisition points;

20. Support data multi-path forwarding and third-party platform access. Generally speaking, neuromorphic computing refers to borrowing the basic rules of information processing in the brain, making fundamental changes to existing computing systems and systems at multiple levels such as hardware implementation and software algorithms, in order to achieve significant improvements in computing energy consumption, computing power, and computing efficiency. The significant development of communication and computer technology in the past few decades has brought about an information revolution, but existing computing systems still face two serious development bottlenecks: first, high system energy consumption, and second, insufficient processing ability for cognitive tasks that the human brain can easily handle (such as understanding language and complex scenes), making it difficult to support high-level intelligence. The obvious advantages of the brain in these two aspects make borrowing from the brain a very promising direction. Brain like computing is a highly interdisciplinary and integrated field of life sciences, especially neuroscience and information technology. Its technical connotation includes a deep understanding of the principles of brain information processing, the development of new processors, algorithms, and system integration architectures based on this, and their application in a wide range of fields such as next-generation artificial intelligence, big data processing, and human-computer interaction. Brain like computing technology is expected to enable artificial information processing systems to generate intelligence comparable to the human brain with very low energy consumption. Many people believe that the substantial progress in this direction may truly open the prelude to the intelligent revolution, bringing profound changes to social production and life. The research on brain computing can be roughly divided into three aspects: neuroscience research, especially the research on the basic principles of brain information processing, the research on brain computing devices (hardware), and the research on brain learning and processing algorithms (software). In the field of neuroscience, there has been very rapid development in the past few decades, especially in the past decade or so. There is now a wealth of knowledge accumulated about the working principles of the brain, which provides an important biological foundation for the development of neuromorphic computing. The human brain is a complex network composed of nearly 100 billion neurons through hundreds of trillions of contact sites (synapses). The material basis for the realization of various brain functions such as sensation, movement, and cognition is the orderly transmission and processing of information within this vast network. Through the efforts of several generations of neuroscientists, there is now a considerable understanding of the structure and function of individual neurons. But there are still many problems to be solved regarding how neurons with relatively simple functions can be organized through networks to form the most efficient information processing system we know today. The brain network is manifested at the microscopic level as connections made up of neural synapses, at the mesoscopic level as connections made up of individual neurons, and at the macroscopic level as connections made up of brain regions and subregions. The information processing carried out on brain networks of different scales has important differences and is closely interconnected, forming a unified whole. At present, the research hotspots in neuroscience mainly focus on analyzing the structure of brain networks at the above-mentioned levels, observing the activity of brain networks, and ultimately elucidating the functions of brain networks, namely the mechanisms of information storage, transmission, and processing. To achieve this goal, the key technologies that need to be broken through are precise and rapid determination of brain network structure, large-scale detection and regulation of brain network activity, and efficient analysis of these massive data. In addition, it is urgent to establish appropriate models and theories under the constraints of experimental data to form a complete understanding of brain information processing. The original intention of researching neuromorphic computing devices is to significantly reduce power consumption without affecting performance, or to greatly improve speed under similar power consumption. Although modern computers have astonishing computing power and speed, they are accompanied by high energy consumption. The power consumption of large computers often exceeds the megawatt level