Development of architectural lighting simulation software

Development of Architectural Lighting Simulation Software of Journal of Guangxi University Lu Yimin Huang Xianfeng 21. School of Electrical Engineering, Guangxi University, Nanning, Guangxi 53,04; 2. School of Civil Engineering and Architecture, Guangxi University, Simulation Software for Architectural Lighting, Nanning 530004, Guangxi. This paper discusses the development and implementation of the software.

Intelligent lighting refers to the application of artificial intelligence methods in architectural lighting and design. The areas covered by intelligent lighting include the intelligentization of lighting equipment and its intelligent control management; the intelligentization of the light environment; the intelligentization of lighting calculation and design. The content discussed in this paper is how to use artificial neural networks to realize the intelligent design of architectural lighting. In the architectural lighting design should be based on the people-oriented design concept, according to visual comfort evaluation indicators, to create a pleasant light environment, so that people can be engaged in a variety of visual work and visually efficient and enjoyable indoors. The artificial neural network technology is developed into a real chat section, which maps the functional relationship between the input feature quantity and the output feature quantity of architectural lighting in violation of thermal engineering architectural lighting and architectural acoustics; Develop architectural lighting design simulation software. Here, a layer 8 network with a hidden layer is selected to solve the lighting design. The number of inputs and the number of outputs of the network are defined by external specific questions.

1 Development of architectural lighting design simulation software 1. Selection of development tools is used for lighting design 8. The network must be realized by means of a large-scale ten-calculation small energy, involving a large amount of ten, large algorithm, 1 Seven necessary to develop external imitation software, laying theory. The bridge between the real. The neural network toolbox 2, which is also developed by the company, includes six kinds of algorithms and functions and examples. The user can use the functions provided by the toolbox to train and simulate the neural network. An excellent platform for neural network simulation. Several people 8 software systems have been widely used in many fields, and people are paying more and more attention to them.

1.2 The function of the simulation software and the function of the overall frame simulation software include two simulations performed by the network trained by the developer. As long as the lighting design requirements are provided as input feature quantities to the network, the simulation calculation results can be obtained.

Simulate on a user-trained network. First, the user trains the network according to his own requirements, determines the structure of the network, and selects the hidden layer network to provide input. Simulation results can also be obtained.

The simulation software adopts a top-down, step-by-step refined modularized programming method. The simulation software program box is as follows: the neural network interface includes 8 network structure simulation process, the error curve is displayed, the simulation result is numerically necessary, the user control element object is also dry, the button slider is used, the 1 input edit box, etc. . When the user calls the mouse to operate on the control meta object, the simulation program will be responded and a predetermined function subroutine will be executed, such as pressing the training simulation button to perform training simulation on the 6 network, and moving the slider to change the hidden layer nerve. The value of the number of elements is entered in the user input edit box, and the input of the trained architectural illumination sample set or the illumination feature quantity to be simulated is input.

1.3 Shape types and relationships and implementation of various operations. Tree relationship of 1 person 8 objects 2 computer screen Jiang coordinate system user control meta object user interface menu line segment area piece surface like text source 2 algorithm program flow elbow 18 neural network toolbox summarizes the existing neural network comparison The mature results, the network models involved are perceptron linear network 8 network radial basis function network self-organizing network and feedback network. For various network models, the Neural Network Toolbox integrates multiple learning algorithms and provides feature-rich tool functions that users can use directly to achieve their goals.

Take 1 is 15 network sequence process 勹 0 + 呲呲 绨 绨 4 呲 岂 岂 岂, broken. Merchants dare to blame 1 葚 葚 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 , to the sense of 32 oil 铷, 4 雒 颉 秦 Qin 5 counter 砘鲥 砘鲥 Zheng series, 洚 谦 赛! Zhejiang. , 83, 澧 6 6 ball 砘 50 50 1 floating, sister; painting to reduce glare. More importantly, this simulation software has the ability of memory and self-learning. Based on the accumulation of a large amount of design experience, it can continuously improve the future lighting design to create a comfortable indoor space for people's visibility. Keep and improve. At present, the intelligent design of the hall-like lighting design can be realized by using the 6 network model and the programming of a few people. But there are still some aspects to be improved. Improve the 8 algorithm or use other artificial neural network models to improve the convergence speed of the iteration.

Progress to enrich and improve the functionality of the simulation software, in order to achieve intelligent design of partition-like lighting and accent lighting in large halls.

In addition to the application of architectural lighting calculation and design, artificial neural network simulation can also be widely used in lighting load calculation lighting circuit design, building lighting design, building thermal environment, indoor sound quality design, air conditioning design and communication design, and then realize the intelligent design of building technology. .

Lou Shuntian, Shi Yang. System analysis and design neural network based on elbow 1 person 6 . Xi'an University of Electronic Science and Technology Press, 1998.

Lu Yimin. Artificial neural network, the development and research of human 1 courseware. Computing Technology and Automation, 2000 Supplement, 4650.

Gao Junbin. Ding 165.0 language and programming. Wuhan Huazhong University of Science and Technology Press, 1998.

Chen Zhonglin, Wang Aiying, Guo Ping. Research on Intelligent Lighting Design Methodology 1. Journal of Chongqing Jianzhu University, 1999, 86768.

Responsible editor Zhang Xiaoyun

LED Advertising Display


Product categories of LED Advertising Display includes MC Pro series Door Led Screen, MC Xs series Cross LED Display, MC Smart series Double sided Led Screen, MC Wall Wall Hanging Led Display,MC Air Showcase LED Display


We are specialized manufacturer for small LED Advertising Display, mature supply chains are available to have competitive price. small LED Advertising Display is widely used in the door led screen industry, it gradually not only becomes the mainstream of the development now, but also city`s economic development coordinates. 


We insist to offer novelty and unique appearance, adjustable size LED screen and will be indicate to supply customers a complete set of LED display solutions.


Display a Visual Future!!!

LED Advertising Display

LED Advertising Display,LED Display Screen,LED Advertising Screen,LED Display

Shenzhen Macion Optoelectronics Technology Co.,Ltd. , https://www.macion-led.com

Posted on