This paper introduces the multivariable predictive control algorithm and its application in the control of environmental test equipment in detail. Because the temperature and humidity control system of environmental test equipment has a large time lag and there is a serious coupling phenomenon between the systems, the conventional PID control can not achieve satisfactory control effect. Aiming at this system, this paper uses multivariable predictive control algorithm to control and simulate it.

Predictive control algorithm is a control algorithm based on system input and output description, and its three basic principles are predictive model, rolling optimization and feedback correction. It chooses the unit step response as its "prediction model". The algorithm can not only simplify the modeling process, but also obtain better control effect and decoupling effect by selecting appropriate design parameters.

This paper briefly introduces the environmental test equipment and explains the problems existing in the control. Then the multivariable predictive control algorithm is derived in detail, including multivariable self-balanced system predictive control algorithm and multivariable non-self-balanced system predictive control algorithm; Then the modeling process of the system and the corresponding system model are given. On this basis, the multivariable predictive control algorithm is used to control and simulate the environmental test equipment, and the simulation results are compared.

The simulation results show that the multivariable predictive control algorithm can achieve more satisfactory results than the conventional PID control for multivariable systems with similar characteristics to the temperature and humidity control system of environmental test equipment.

Keywords: multivariable system; Predictive control; Environmental test equipment

Chinese abstract * * * 100-300 words, keywords 3-7 words.

Chinese abstracts and keywords take up one page.

All English is in Times New Roman font.

Abstract font number three, bold, middle line, top and bottom line.

The text is in small font No.4, and the spacing is fixed at 20 points.

In this paper, the multivariable predictive control algorithm and its application in environmental test device control are introduced in detail. The temperature and humidity control system of environmental test device has the characteristics of large time delay and strong coupling. Therefore, the conventional PID control effect is not ideal. In this case, the temperature and humidity control of environmental test device based on multivariable predictive control algorithm is simulated.

Predictive control algorithm is a control algorithm based on system input and output description. Its three basic principles are predictive model, rolling optimization and feedback correction. The unit step response is selected as the prediction model, which simplifies the modeling process. In addition, by choosing appropriate parameters, good control and decoupling effects can be obtained.

This paper briefly introduces the environmental test device and points out the existing problems. Then the multivariable predictive control algorithm is introduced in detail, including multivariable automatic balance system predictive control algorithm and multivariable automatic imbalance system predictive control algorithm. Secondly, the system modeling process and the corresponding system model are proposed. In addition, the multivariable predictive control algorithm is applied to the temperature and humidity control system of environmental test equipment. Finally, the simulation results are compared.

The simulation results show that the multivariable predictive control algorithm can be used in multivariable systems such as temperature and humidity control system of environmental test device, and its control effect is more satisfactory than that of conventional PID control.

Keywords: multivariable system; Predictive control; Environmental test equipment

English abstracts and keywords should be translated into Chinese abstracts and keywords.

English abstracts and keywords take up one page.

Directory sample, automatically generated by word.

catalogue

Chapter 1 Thread Theory 1

1. 1 Introduction 1

1.2 application and development of digital image technology

1.3 question 3

1.4 Arrangement of chapters in the paper 4

Chapter II Digital Image Processing Methods and Research 5 pages

2. 1 gray histogram 5

2. 1. 1 definition 5

2. Properties and uses of1.2 histogram 5

2.2 Geometric transformation

2.2. 1 space transformation 8

2.2.2 Gray interpolation

2.2.3 Application of Geometric Operation 10

2.3 spatial filtering enhancement 10

2.3. 1 spatial filtering principle 10

2.3.2 Laplace operator 1 1

2.3.3 median filtering 12

2.4 image segmentation processing 13

2.4. 1 histogram threshold binary segmentation 14

2.4.2 Optimal threshold segmentation of histogram 14

2.4.3 Regional growth 16

Chapter III Image Processing Software Design 18

3. 1 Select image processing software development tool 18

Structure 3. 1. 1 BMP image format 18

3. 1.2 Select software development tools 19

3.2 brief introduction of ean-13 code 20

3.2. Structure of1ean-13 barcode 20

3.2.2 Coding mode of barcode 2 1

3. 1 System Interface Design 22

The fourth chapter barcode image test 24

4. 1 Main methods of barcode image processing 24

4.2 Bar code image test results 25

Chapter V Summary and Prospect 28

Reference 29

When the prior probabilities are equal, that is, when

(2.33)

It's just the average of the two.

As can be seen from the above analysis, as long as the sum is known and normal, it is easy to calculate its optimal threshold t.

The parameters of the actual density function are usually estimated by fitting method. E.g., minimum mean square error fitting estimation, to account for the parameters and minimize the mean square error of fitting. For example, suppose the density of the ideal distribution is normal, the histogram of the actual image is 0, and the fitting error in a discrete way is 0.

(2.34)

Where n is the abscissa of the histogram. Usually, it is difficult to find several parameters of the density function by this fitting, and the numerical solution can only be obtained by computer, but if it is a normal distribution, only two parameters are needed: mean and standard deviation.

2.4.3 Regional growth

Region growing is a typical continuous region segmentation technology and a very important image segmentation method in computer vision research in the field of artificial intelligence. Its main idea is to gather pixels satisfying some similarity judgment around pre-selected seed points to form an area. In concrete processing, it begins with dividing an image into many small areas, which are generally small neighborhoods or even single pixels. Then, by defining appropriate membership rules in the region to check the surrounding pixels, the pixels that meet the above membership rules are merged, otherwise they are discarded, and after several iterations, the region to be segmented can finally be formed. The "internal membership rule" mentioned here can be determined according to many factors such as gray level characteristics, texture characteristics and color characteristics of the image. It can be seen from this passage that the key to the success of regional growth lies in choosing appropriate internal membership rules (growth standards).

For the growth criterion based on image gray features, the following process can be used to describe its regional growth process, as shown in Figure 2.6.

Figure 2. 6 regional growth flow chart

Chapter III Design of Image Processing Software

3. 1 Selection of development tools for image processing software

3. Structure of1.1BMP image format

There are many formats for storing digital images, such as BMP, GIF, JPEG, TIFF, etc. BMP is the most commonly used in digital image processing, and the pictures collected in this subject are also stored in BMP format. To deal with this format of the picture, we must first understand its file structure.

Brief introduction of (1)BMP file format

BMP (bitmap file) graphic file is a graphic file format adopted by Windows. All image processing software running in Windows environment supports BMP image file format. All image drawing operations in Windows system are based on BMP. Before Windows 3.0, the BMP bitmap file format was related to the display device, so this BMP image file format was called DDB (Device Related Bitmap) file format. The BMP image file after Windows 3.0 has nothing to do with the display device, so this BMP image file format is called DIB (Device Independent Bitmap) format, in order to enable Windows to display the stored images on any type of display device. The default file extension of bmp bitmap file is BMP or BMP (sometimes required. DIB or. RLE is the extension).

(2) Composition of 2)BMP file

BMP file consists of bitmap file header, bitmap information header, color table and graphic data. Its form is shown in Table 3. 1.

Table 3. 1 BMP bitmap structure

Composite structure name symbol of bitmap file

Bitmap file header

Bitmap information title Bitmap information title bmih

Color table RGB four colors []

Graphics data byte bit []

3. 1.2 Selection of software development tools

( 1)Win32 API

Microsoft Win32 API (Application Programming Interface) is an application programming interface of Windows, including window information, window management function, graphics device interface function, system service function, application program resources and so on. Win32 API is the foundation of Microsoft's 32-bit Windows operating system and all 32-bit Windows applications.

These programs all run on Win32 API, and their functions are provided by the system's dynamic link library.

(2)Visual C++

Visual C++ is a visual programming product produced by Microsoft, which is object-oriented, closely integrated with Windows API, rich in technical resources and powerful in auxiliary tools. Since its birth, Visual C++ has been one of the most important application development systems in Windows environment. Visual C++ is not only an integrated development environment of C++ language, but also closely related to Win32. Therefore, Visual C++ can be used to develop various applications, from the bottom software to the top software directly facing users. Visual C++ is a visual programming environment with friendly interface and easy operation for programmers.

Visual C++ can make full use of the advantages of MFC. There are many basic library classes in MFC, especially some in MFC, which can be used to write various Windows applications, saving a lot of repetitive work time and shortening the development cycle of applications. Using MFC's basic class library will get twice the result with half the effort when developing applications.

Visual C++ has the following characteristics:

Simplicity: Visual C++ provides a series of wizard tools, such as MFC class library, ATL template class, AppWizard and ClassWizard, to help users build their own applications quickly, which greatly simplifies the design of applications. Using these technologies, developers can develop Windows applications with little or no code.

Flexibility: The development environment provided by Visual C++ enables developers to design the interface and functions of applications according to their own needs. Moreover, Visual C++ provides a wealth of class libraries and methods, so that developers can choose according to their own application characteristics.

Extensibility: Visual C++ provides OLE technology and ActiveX technology, which can enhance the capabilities of applications. Using OLE technology and ActiveX technology, developers can use various components and controls provided by Visual C++ and components provided by third-party developers to create their own programs, thus realizing the componentization of applications. Using this technology can make the application program have good scalability.

(3)MFC

MFC (Microsoft Basic Class) is a set of basic classes developed by Microsoft in C++ language.

Ku。 It is complicated to program directly with Win32 API, and Win32 API is not object-oriented. MFC encapsulates most of the contents of Win32 API and provides an application framework to simplify and standardize the design of Windows programs. MFC is an important part of Visual C++, and it is integrated with it in the most ideal way. It mainly includes the following parts: encapsulation of Win32 API, application framework, OLE support, database support, general classes and so on.

3.2 brief introduction of ean-13 code

The bar code printed on commodity packaging, which people see every day, has been widely used in industry, commerce, national defense, transportation, finance, medical and health care, post and telecommunications, office automation and other fields since it came out in the early 1970s. According to different classification methods and different coding rules, bar codes can be divided into many kinds. There are as many as 250 kinds of bar codes in use in the world. This chapter takes the standard version of EAN bar code EAN- 13 as an example to explain the development method of EAN bar code image recognition software based on digital image processing technology.

EAN code is a commodity bar code popularized and applied by ean international all over the world. It is a fixed-length pure digital bar code, and its character set is numbers 0 ~ 9. It consists of prefix code, manufacturer identification code, commodity item code and check code. Prefix code is a code used by international EAN organizations to identify member organizations, which is 690 ~ 695 in China; The supplier identification code is the code assigned to the supplier by the EAN member organization according to the EAN prefix code; The commodity item code is coded by the manufacturer; The check code is used to check the correctness of the previous 12 or 7-bit code.

3.2. 1 EAN- 13 barcode structure

EAN- 13 code is coded by "modular combination method". Its symbol structure consists of eight parts: left blank area, start symbol, left data symbol, middle separator, right data symbol, check symbol, terminator and right blank area, as shown in Table 3.2. Size: 37.29mm×26.26mm；; Bar code: 31.35mm; Initiator/separator/terminator: 24.50 mm; The amplification factor ranges from 0.80 to 2.00; The interval is 0.05.

Table 3. 2 EAN- 13 code structure

left

To the left of the space start character

Data symbol middle

Right side of separator

Data symbol check terminator

frontage

blank space

9.

3 modules

42 modules

5 modules

35 modules

7 modules

3 modules

9 modules

package

The code represented by EAN- 13 code consists of 13 digits, and its structure is as follows:

Structure 1:

x 13x 12x 1 1x 10x 9 x 8 x 7 x 6 x 5 x 4 x 3 x 1

Where: X 13 ~ X 1 1 is the prefix code representing the country or region code; X 10 ~ X7 is the manufacturer code; X6 ~ x2 are the codes of commodities; X 1 is the check code.

Structure 2:

x 13x 12x 1 1x 10x 9 x 8 x 7 x 6 x 5 x 4 x 3 x 1

Where: X 13 ~ X 1 1 is the prefix code representing the country or region code; X 10 ~ X6 is the manufacturer code; X5 ~ x2 are commodity codes; X 1 is the check code.

In China, when x 1 3x12x11is 690 and 69 1, the code structure is the same as the structure1; When x13x12x11is 692.

When its code structure is the same as structure 2.

See table 3.3 for the coding rules of EAN bar codes:

Initiator:101; Middle separator: 01010; Terminator: 10 1.

"0" and "1" in A, B and C respectively represent "empty" and "bar" of a module width.

Table 3. Three coding rules of EAN bar code

Left side of data symbol

Right side of data symbol

Data symbol

B.C.

0 000 1 10 1 0 100 1 1 1 1 1 100 10

1 00 1 100 1 0 1 100 1 1 1 100 1 10

2 00 100 1 1 00 1 10 1 1 1 10 1 100

3 0 1 1 10 1 0 10000 1 10000 10

4 0 1000 1 1 00 1 1 10 1 10 1 1 100

5 0 1 1000 1 0 1 1 100 1 100 1 1 10

6 0 10 1 1 1 1 000 10 1 10 10000

7 0 1 1 10 1 1 00 1000 1 1000 100

8 0 1 10 1 1 1 000 100 1 100 1000

9 000 10 1 1 00 10 1 1 1 1 1 10 100

3.2.2 Coding method of bar code

The coding method of bar code refers to the coding rules of empty white bars in bar code and the setting of binary logic representation. As we all know, computer equipment can only read binary data (data has only two logical representations of "0" and "1"). Bar code symbol, as a graphic symbol of photoelectric scanning information provided for computer information processing, should also meet the requirements of computer binary. The coding method of bar code is to represent different binary data by designing the arrangement and combination of bars and spaces in bar code. Generally speaking, there are two kinds of bar codes: module combination and width adjustment.

Module combination method refers to bar code symbols, in which bars and spaces are composed of modules with standard width. The standard width bar represents binary "1", and the standard empty module represents binary "0". The standard width of bar code module is 0.33mm, and a character consists of two bars and two spaces, and each bar or space consists of 1 ~ 4 standard width modules.

The width adjustment method refers to the bar code, in which the width of the bar is different, and the binary "1" is represented by the wide unit, and the binary "0" is represented by the narrow unit, and the ratio of the width to the width unit is generally controlled between 2 and 3.

3. 1 system interface design

The basic functions of image processing software in this paper include reading images, saving images and processing images. Figure 3. 1 shows the interface of the image processing software.

Figure 3. 1 software main interface

The software design flow chart is shown in Figure 3.2.

Figure 3. 2 program design flow chart

Chapter IV Barcode Image Testing

4. 1 Main methods of barcode image processing

(1)256 color bitmap is converted into gray image.

The application of gray processing in point processing provides a prerequisite for realizing threshold transformation of digital images. To convert a 256-color bitmap into a grayscale image, the corresponding grayscale value of each color must be calculated first. The corresponding relationship between grayscale and RGB colors is as follows:

y = 0.299 r+0.587g+0. 1 14B(4. 1)

In this way, according to the above formula, we can easily convert the 256-color palette into a gray palette. Because the palette of gray-scale images is generally arranged in the order of increasing gray-scale, we must also adjust each pixel value of the image (that is, the index value of the palette color). In actual programming, we only need to define a mapping table bMap[256] (a one-dimensional array with a length of 256, which stores the gray value corresponding to each color in the 256-color palette), and replace each pixel value p (that is, the color index value in the original 256-color palette) with bMap[p].

(2) threshold transformation of gray scale

Using the threshold transformation theory in point operation, the gray image is transformed into binary image to prepare for image analysis. Gray threshold transformation can transform gray image into black and white binary image. Its operation is to specify a threshold by the user. If the gray value of a pixel in the image is less than the threshold, the gray value of the pixel is set to 0, otherwise the gray value is set to 255.

(3) Median filtering

The spatial filtering method in transform domain method is used to denoise the image. Median filtering is a nonlinear signal.

The processing method and the corresponding filter are also nonlinear filters. Median filtering generally adopts a sliding window with an odd number of points, and replaces the gray value of the specified point (usually the center point of the window) with the median value of the gray value of each point in the window. For odd-numbered elements, the median value refers to the middle value sorted by size, and for even-numbered elements, the median value refers to the average of the gray values of the two middle elements after sorting.

(4) Vertical projection

Using vertical projection method in image analysis to reconstruct binary image provides a prerequisite for bar code recognition. Vertical projection is the transformation of black and white binary images by projection. The height of the black line in the transformed image represents the number of black spots in the column.

(5) Geometric operation

Geometric operations can change the spatial relationship between objects in an image. An important application of geometric operation is to eliminate the geometric distortion of digital images caused by cameras. Geometric correction has proved to be very important when quantitative spatial measurement data need to be obtained from digital images. In addition, some imaging systems use non-rectangular pixel coordinates. When observing these images with ordinary display equipment, we must first straighten them, that is, convert them into rectangular pixel coordinates.

4.2 Bar code image test results

The processing object of this software is 256-color BMP bitmap of EAN- 13 code. Using the methods of digital image processing technology, such as gray processing, threshold segmentation, spatial filtering, region growth and projection, the noisy barcode image is processed, and the results are as follows:

Figure 4. 1 original barcode Tu Tu 4. 2 gray window transformation

Figure 4. 3 original bar code histogram Figure 4. 4 Gray Window Transform Histogram

Figure 4. 5 gray histogram specification interface Figure 4. 6 gray histogram specification histogram

Figure 4. 7 median filtering interface

Figure 4. 8 Regional growth Figure 4. 9 threshold area elimination

Figure 4. 10 vertical projection

From the above processing results, it can be seen that the original barcode image is transformed into a projection image after gray scale transformation, median filtering, binarization and small area threshold elimination. Whether the barcode can be read out by image pattern recognition in the next step needs further study.

Chapter V Summary and Prospect

Digital image processing technology originated in the 1920s. At that time, due to the limitation of technical means, the development of image processing technology was slow. It was not until the emergence of the third generation computer that digital image processing was rapidly developed and widely used. Today, there are almost no technical fields unrelated to digital image processing.

This paper mainly studies the related knowledge of digital image processing, and then realizes the image processing algorithm through Visual C++. All the algorithms mentioned in this paper are processed and conclusions are drawn. The work done is as follows:

The gray processing in (1) point processing method provides a prerequisite for realizing the threshold transformation of digital images.

(2) Using the spatial filtering method in the transform domain method to denoise the image.

(3) Using the threshold transformation theory in point operation, the gray image is transformed into a binary image to prepare for image analysis.

(4) Using the vertical projection method in image analysis to realize the reconstruction of binary image, which provides a prerequisite for barcode recognition.

In the last chapter, the results of various algorithms are given. The results show that the noisy bar code can be processed into noise-free bar code through digital image processing.

The application fields of digital image processing technology are various, which can be used not only for image processing described in this paper, but also for pattern recognition and machine vision. In recent years, the image processing methods developed on the basis of morphology and topology have brought a new situation to the field of image processing, and it is believed that the application of image processing will be more extensive in the future.

refer to

Ruan Qiuqi. Digital image processing [M]. Beijing: Electronic Industry Press, 200 1.

Yellow,,,. Digital image processing and compression coding technology [M]. Chengdu: University of Science and Technology Press, 2000.

Look proud. Computer image processing. Beijing: Tsinghua University Publishing House, 2000.

Hu. Data Structure-A Guide to Algorithm Design [M]. Beijing: Tsinghua University Publishing House, 1999.

Huang weitong. Object-oriented and visual programming [M]. Beijing: Tsinghua University Publishing House, 200 1.

Xia. Digital image processing [M]. Nanjing: Southeast University Press, 1999.

Fei Zhenyuan Bar code technology and application [M]. Shanghai: Shanghai Science and Technology Literature Publishing House, 1992.

Li Jinzhe. Automatic bar code recognition technology [M]. Beijing: National Defense Industry Press, 199 1.

He Bin. Digital image processing [M]. Beijing: People's Posts and Telecommunications Publishing House, 200 1.

[10] Li Changjiang. C++ user manual [M]. Beijing: Electronic Industry Press, 1995.

Zhang Chunlin and Qin Xi. Visual C++ 6.0。 Practical programming technology [M]. Beijing: China Water Resources and Hydropower Press, 1999.

[12] Hu data structure-algorithm design guide [M]. Beijing: Tsinghua University Publishing House, 1999.

[13] Kenneth R. Casselman, Zhu Zhigang et al. Digital image processing [M]. Beijing: Electronic Industry Press, 1998.

Davis Chapman Visual C++6.0 [M]. Beijing: Tsinghua University Publishing House, 1999.

Richard c le necker. Visual C++5 PowerToolkit [M]。 Beijing: Machinery Industry Press, 1999.