Improved space mapping optimization algorithm
A space mapping and optimization algorithm technology, applied in the field of improved space mapping optimization algorithm, can solve the problems of single training data, high cost, and inaccuracy of mapping relationship, and achieve the effect of accurate training process and time saving
Inactive Publication Date: 2018-01-26
TIANJIN UNIV
View PDF0 Cites 1 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, this method needs to solve the fine model 2n+1 times at the beginning of training, which is time-consuming and expensive whether it is obtained through electromagnetic simulation or actual measurement
Therefore, the current popular spatial mapping algorithm simplifies the process to a great extent, that is, only a set of optimal solutions of the coarse model is used for training [3], which greatly saves time and cost, but the obtained mapping relationship is often due to training The data is single and not accurate enough
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0033] In order to illustrate the technical solution of the present invention more clearly, the present invention will be further described below in conjunction with the accompanying drawings. For those skilled in the art, other drawings can also be obtained based on these drawings without creative effort.
[0034] In order to solve the above technical problems, the present invention improves the original neural network space mapping optimization algorithm, and the algorithm flow chart is as follows figure 1 shown. The improved space mapping optimization algorithm of the present invention, concrete process is as follows:
[0035] Step 1, get the rough model input x of the optimal output response according to the design goal c * :
[0036]
[0037] Among them, R c (x c ) is the response of the coarse model.
[0038] Step 2, input the rough model into x c * Directly as input to the fine model:
[0039]
[0040] Step 3, calculate the response of the fine model D...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses an improved space mapping optimization algorithm, which includes obtaining a rough model input xc* of an optimal output response according to a design objective; taking the xc*directly as an input of a fine model; calculating a fine model response, and judging whether the fine model response satisfies the following shown in the description; if so, obtaining an optimal finemodel response; otherwise, establishing a neural network space mapping relation as shown in the description; training a neural network mapping function P so as to satisfy the following as shown in the description; optimizing a fine model parameter by using the training P to obtain the optimal fine model parameter, taking the optimal fine model parameter as a next input of the fine model, and repeating circularly until the optimal fine model response is obtained. The improved space mapping optimization algorithm in the invention reduces the cost of the traditional space mapping and improves the simulation accuracy of the current popular space mapping.
Description
technical field [0001] The invention belongs to the field of microwave engineering, specifically relates to the electromagnetic simulation optimization of radio frequency and microwave circuits, and more specifically relates to an improved space mapping optimization algorithm. Background technique [0002] Space Mapping is a well-recognized optimization method in the field of microwave engineering, which can combine the computational efficiency of the coarse model with the accuracy of the fine model. Coarse models are usually empirical formulas or equivalent circuit models, with very high calculation efficiency, but often the range of parameters is limited, so when the effective range is exceeded, the simulation results will be inaccurate; while the fine model can be provided by electromagnetic simulation software, Or it can be measured directly, so the accuracy is very high, but the cost is high (such as CPU-intensive simulation). The spatial mapping can establish a mathem...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): G06F17/50
Inventor 马建国王彩霞傅海鹏
Owner TIANJIN UNIV