Office Online  
    Office Online
    Manuscript Tracking
    Editor Work
    Office Work
    Editor-in-chief
Journal Online
    Current Issue
    Advanced Search
    Archive
    Read Articles
    Download Articles
    Email Alert
Traffic
Quick Search  
  Adv Search
2021 Vol. 42, No. 5
Published: 2021-05-28
 
545 Blind Reconstruction Method of AFM Image Based on Conditional Generative Adversarial Network
HU Jia-cheng,YAN Di-xin,CAO Cong,SHI Yu-shu,ZHANG Shu,LI Dong-sheng
DOI: 10.3969/j.issn.1000-1158.2021.05.01
To solve the problem caused by the tip broadening effect in the imaging process of atomic force microscope(AFM), a blind reconstruction method of AFM image based on the condition generating adversarial network (CGAN) is proposed.First, on the basis of pix2pixHD network, the simulation sample data are trained by global generation network, and AFM measurement data are trained by local lifting network. Finally, feature matching loss function is used to improve the lateral resolution of grid edge.The experimental results show that when performing the blind reconstruction for the measurement image of the one-dimensional rectangular grid with line width being 8μm under AFM, the standard deviation of the reconstructed image is 0.33×0.45μm, which has a high imaging resolution and is conducive to improving the accuracy of the measurement of the AFM image one-dimensional grid.
2021 Vol. 42 (5): 545-551 [Abstract] ( 266 ) HTML (1 KB)  PDF (2545 KB)  ( 167 )
552 Calibration Algorithm and Experimental Research of Cooperative Robot Based on Laser Tracker
CHEN Xiang-jun,ZUNONG Gulinaer,XUE Zi,LI Da-chao,BAN Zhao,REN Guo-ying
DOI: 10.3969/j.issn.1000-1158.2021.05.02
In order to improve the absolute positioning accuracy of the robot end,a method for calibrating the geometric parameters of the robot based on laser tracker measurement is proposed to identify and compensate the parameter errors of the cooperative robot.In order to avoid singularity when the two axes of the robot arm are parallel,MDH parameter method is used to establish the error model. In order to define the measurement data on the same coordinate axis,the robot tool coordinate transformation is combined (the translational transformation of the target ball′s center point relative to the robot′s terminal coordinate system),and the levenberg marquardt (LM) method is used to identify the model parameter error of the robot.Through the simulation,experimental calculation and calibration of JAKA cooperative robot,the average error,the standard deviation and the maximum error of the robot are reduced respectively by 70.58%,56.76% 57.44%.The results show that the calibration algorithm can effectively identify the model parameter errors of the robot,and compensate the model parameters of the robot to improve the absolute positioning accuracy.
2021 Vol. 42 (5): 552-557 [Abstract] ( 281 ) HTML (1 KB)  PDF (1565 KB)  ( 521 )
558 Study on the Influence of the Calculation Method on the Accuracy of the Multilateral Coordinate Measurement System
SUN Wei,MIAO Dong-jing,LI Jian-shuang,YAO Yan,ZHONG Wen,HE Ming-zhao,LI Lian-fu
DOI: 10.3969/j.issn.1000-1158.2021.05.03
In order to study the influence of the calculation method on the accuracy of the multilateral coordinate measurement system,the multilateral coordinate measurement model was established,and the differences between the two methods of coordinate calculation were analyzed.The simulation experiment of the two methods was carried out for the typical multilateral coordinate measurement system of four stations.The simulation results show that the pre-accurate calibration of system parameters can effectively improve the measurement accuracy,which is improved by 69.5%,64.6% and 46.3% respectively.An experiment was conducted to verify the accuracy of coordinate measurement.The experimental results show that,compared with the synchronous calculation method,the average measurement error of the three groups of experiments after the pre-accurate calibration of system parameters decrease from 203.0μm to 23.8μm,and the average correlation coefficient between the measurement error and the measurement distance of the three groups of experiments is reduced from 99.8% to 37.8%,which verify the simulation results.
2021 Vol. 42 (5): 558-563 [Abstract] ( 211 ) HTML (1 KB)  PDF (1843 KB)  ( 318 )
564 Study on Thickness Evaluation Method of Sea SurfaceEven Thick Oil Film Area Edge Based on LIF
XIE Bei-bei,KONG Ling-fu,KONG De-ming,ZHANG Xiao-dan,KONG De-han,YUAN Li
DOI: 10.3969/j.issn.1000-1158.2021.05.04
Oil film thickness is an important index for quantitative analysis of sea surface oil spill pollution. Based on the morphological characteristics of the even and uneven thick oil film areas in the sea surface oil spill, the determination method was given based on the signal characteristics of laser-induced fluorescence (LIF) detecting oil film with different thickness, and the thickness evaluation method of the edge of even thick oil film area was proposed. The oil film thickness was estimated according to the fluorescence intensity corresponding to the thickness of the oil film area edge. The thickness change of the even thick oil film edge was observed, and the thickness of the even thick oil film was evaluated by LIF experimental system. The results show that the relative deviation of this method is about ± 10% for the oil film thickness greater than 20 μm.
2021 Vol. 42 (5): 564-569 [Abstract] ( 281 ) HTML (1 KB)  PDF (2065 KB)  ( 174 )
570 Shadow Detection Method Combining Multi-Scale and Dense Feature Map Fusion
ZHANG Shi-hui,ZHANG Xiao-wei,LI He,ZHANG Xiao-xiao,NIU Jing-chun,CHEN Qi
DOI: 10.3969/j.issn.1000-1158.2021.05.05
In order to improve the accuracy of shadow detection in the image, a shadow detection method utilizing deep neural network is proposed. Firstly, a dense feature map fusion structure is proposed to fuse the feature maps generated by different convolutional layers. Secondly, a serial-parallel dilated convolution structure is designed to extract the multi-scale feature in the original image aiming to the scale variant phenomena in shadow detection task. Finally, combining the dense feature map fusion structure and serial-parallel dilated convolution structure, an end-to-end dilated dense fusion-unet is constructed to detect shadow. Experimental results demonstrate that the shadow detection results and quantitative evaluation of the proposed method on the SBU and UCF shadow detection datasets outperform the existing representative shadow detection methods, the accuracy on the two datasets increased by 5.8% and 6.5%, and the balance error rate decreased by 2.2% and 0.5%, respectively. The ablation study verifies the structure rationality of the proposed dilated dense fusion-unet.
2021 Vol. 42 (5): 570-576 [Abstract] ( 231 ) HTML (1 KB)  PDF (2941 KB)  ( 275 )
577 The Evaluation of Uncertainty of Valuing System for Heat Sink
SUN Dao-an,Lü Jian,LI Chun-ying,DU Yong-mei,WANG Zhi-xuan, ZHANG Jian-wei,ZHANG Gao
DOI: 10.3969/j.issn.1000-1158.2021.05.06
Heat sink is the characteristic parameter weighing the heat absorption capacity of endothermic hydrocarbon fuels (EHFs). Measuring the heat sink with high efficiency and accuracy is the premise and key technology for the engineering application of EHFs to hypersonic aerocrafts. A high accurate and automated valuing system for heat sink was developed according to the functional modularity design philosophy. On this basis, main uncertainty contributions were analyzed and then the uncertainty of the valuing system for heat sink was evaluated referring to JJF1059 Evaluation and Expression of Uncertainty in Measurement. The results showed that the expanded uncertainty of the valuing system is 2.48% (k=2). The valuing system for heat sink provides a strong support to the metering of the heat absorption capacity of EHFs as well as a basis for the subsequent development of the reference material of heat sink.
2021 Vol. 42 (5): 577-581 [Abstract] ( 216 ) HTML (1 KB)  PDF (827 KB)  ( 202 )
582 Highly Precise Realization of the Indium Freezing Point
WANG Ning,YAN Xiao-ke,ZHANG Ming-yu,HE Pei
DOI: 10.3969/j.issn.1000-1158.2021.05.07
A stainless-steel-cased indium cell was utilized to precisely realize the indium melting and freezing plateaus using a continuous heat flux method. The liquidus temperatures of indium were determined by one linear fitting method, and the differences between the liquidus temperatures and the maximum values of the freezing plateaus were regarded as indicators for evaluating the quality of the indium cell. The results showed that these differences were within 0.27mK. Also, the vertical temperature uniformity within 16cm from the thermometer well bottom did not exceed 13mK. Therefore, the stainless-steel-cased cell can meet the requirements for high-precision temperature dissemination. In addition, the influence of furnance setting temperatures on the duration of the freezing plateau and supercooling were also studied.
2021 Vol. 42 (5): 582-588 [Abstract] ( 250 ) HTML (1 KB)  PDF (2405 KB)  ( 183 )
589 Influence of Self-heating Effect of Standard Platinum ResistanceThermometer on Measurement Results
REN Jian-ping,SUN Jian-ping,LI Ting,HE Jia-rong,ZENG Jia-xu
DOI: 10.3969/j.issn.1000-1158.2021.05.08
In order to study the effect of self-heating on the measurement results of SPRTs, the research was carried out from the fixed-point method and the comparative method. For the fixed-point method, based on the statistics of SPRT verified by the National Institute of Metrology of China in the past three years, the measurement results of SPRT in different temperature regions before and after self-heating correction were compared. The results showed that the effect of self-heating correction on the value traceability of SPRT was more than 1.5mK, and the maximum reached 6mK. For comparative method, a self-heating measurement scheme based on a constant temperature bath was designed and experimented, and the temperature measurement results before and after the self-heating correction were also compared. The results showed that the deviation caused by self-heating correction still reached 1.5mK. Therefore, it is necessary to correct the effect of self-heating in the precision temperature measurement and value traceability of SPRT.
2021 Vol. 42 (5): 589-594 [Abstract] ( 272 ) HTML (1 KB)  PDF (1949 KB)  ( 261 )
595 ECT Image Reconstruction Based on Improved Half-threshold Iterative Algorithm
MA Min,LIU Yi-fei,LIU Ya-nan
DOI: 10.3969/j.issn.1000-1158.2021.05.09
To solve the problem of ill-posedness and underdetermination for the inverse problem of electrical capacitance tomography, the theory of compressed sensing was applied to the imaging process to alleviate its underdetermination. First, the initial signal was sparsed processing, and then the rows of the sensitivity matrix were rearranged based on the Gaussian random matrix, then the singular value decomposition (SVD) was used to obtain the observation matrix with higher column independence. Finally, the half-threshold iterative algorithm based on l1/2 norm was introduced into the ECT imaging process, and the constraint term of l2 norm was added to the penalty function, and solved by the improved semi-threshold iterative algorithm. The simulation experiment showed that the algorithm effectively reduced the image error and took into account the imaging speed, and had good performance in the ECT imaging process.
2021 Vol. 42 (5): 595-602 [Abstract] ( 260 ) HTML (1 KB)  PDF (2184 KB)  ( 234 )
603 Dynamic Calibration Method of a Force Transducer Based on Parameter Identification
JIANG Wen-song,YIN Xiao,LI Hong-yang,LUO Zai,YANG Jun,WANG Zhong-yu
DOI: 10.3969/j.issn.1000-1158.2021.05.10
To evaluate the influence of structural parameters on the time-varying characteristics of a force transducer, a dynamic calibration method is proposed. First, a parametric mathematical model is built according to the structural of the transducer and the transfer function. Second, the model parameters of the transducer are identified from its spectral characteristics by utilizing the least square method. Third, the geometric relations of the transfer function in the frequency domain is built to evaluate the stability of the calibration model based on a frequency analysis method. As well, the suggested method is verified by a drop hammer impact force calibration device. The experimental result shows that the aboved method can identify the structural parameters of the transducer with a small standard deviation and evaluate the relative stability of the calibrated model.
2021 Vol. 42 (5): 603-608 [Abstract] ( 255 ) HTML (1 KB)  PDF (2740 KB)  ( 500 )
609 Error Analysis of Draft Survey Based on Error Transfer Principle
ZHU Jing-lin,ZHU Jin-shan
DOI: 10.3969/j.issn.1000-1158.2021.05.11
Based on the overview of the error correlation theory and the draft survey method, the error propagation law of the draft survey were analyzed. By reading a large number of documents to sort out the systematic errors in the weight measurement process, in many errors, the systematic error caused by the vertical component of the anchor chain and the cable has not been effectively corrected. Based on the JTS144-1-2010 "Port Engineering Load Specification"(hereinafter referred to as specifications) and the existing research results, the lateral and longitudinal component forces of ship wind pressure and water flow force are analyzed and obtained. The horizontal external force of the hull (i.e., the reaction force of the horizontal tension component of the anchor chain) were combined with the mechanical principle and the anchoring state of the ship to establish a calculation model of the vertical component of the anchor tension, which can be used to solve the correction of systematic error caused by the vertical component of anchor chain. The example showed that when the wind speed is 13m/s and the flow rate reaches 4kn (1kn=1.852km/h), only this error has reached 28% of the allowable error of the water gauge.
2021 Vol. 42 (5): 609-614 [Abstract] ( 256 ) HTML (1 KB)  PDF (459 KB)  ( 150 )
615 Bearing Fault Diagnosis of Wind Turbine Based on Multi-wavelet-1D Convolutional LSTM
CHEN Wei-xing,CUI Chao-chen,LI Xiao-jing,ZHAO Hui
DOI: 10.3969/j.issn.1000-1158.2021.05.12
To solve the problem of high false alarm rate of wind turbine bearing fault diagnosis under complex conditions, an end-to-end hybrid deep learning framework is proposed. One-dimensional convolutional recurrent neural network based on multiple wavelet transforms (multi-wavelet-1D Convolutional LSTM, Mw-1DConvLSTM). Firstly, multiple time-frequency maps are obtained by multiple wavelet transforms to fully extract the signal features. Then, an extended LSTM is used to extract different time step information of multi-channel time-frequency maps, and time-space characteristics of time-frequency data are captured. Finally, the fault state is classified by the global pooling layer and the classification layer. The test results show that under complex conditions, Mw-1DConvLSTM can achieve more than 95% fault identification of wind turbine bearing faults.
2021 Vol. 42 (5): 615-622 [Abstract] ( 251 ) HTML (1 KB)  PDF (3019 KB)  ( 579 )
623 DC High Current On-Line Calibration Devicebased on Digital Image Recognition and Optical Fiber Sensor
MEI Guo-jian,WANG Jia-fu,LI Chuan-sheng,LIN Fei-peng,DONG Ping,SHAO Hai-ming,CAI Jin-hui
DOI: 10.3969/j.issn.1000-1158.2021.05.13
In the field of electrolytic metallurgy and electrochemical industry, a large number of Hall DC high current sensor have no reservation calibration interface. Aimed at the practical problem, a DC large current online calibration device based on digital image recognition and fiber current sensor was designed. The system used the optical-fiber current sensor as the standard and collects the display image of the Hall current sensor secondary display instrument in real time by the image sensor. Besides, it performed some pre-processors, such as grayscale, median filtering,binarization, closing operation, character segmentation, etc. on the collected image. By adopting the threading method for image recognition, the sampling value of the Hall current sensor was obtained. Then, after compared with the standard measurement value of the fiber-optic current sensor in real time, the measurement error of the controlled Hall current sensor was calculated. The results of field calibration experiments verified the feasibility of the calibration method and device.
2021 Vol. 42 (5): 623-628 [Abstract] ( 224 ) HTML (1 KB)  PDF (2614 KB)  ( 330 )
629 Motion Imagery Classification Algorithm Research Based on Hybrid Transfer Learning and Application in Brain-computer Interface
DU Yi-hao,LIU Zhao-jun,FU Zi-hao,ZHANG Yuan-yuan,REN Na,CHEN Jie,XIE Ping
DOI: 10.3969/j.issn.1000-1158.2021.05.14
To improve the efficiency and universality of transfer learning in the application of motor imagery brain-computer interface (MI-BCI),a hybrid transfer learning model with integrating the advantages of instance transfer and feature transfer learning methods was built.Firstly,the transfer of the instance level by introducing the principle of sample weight polarization to improve the classical TrAdaBoost algorithm was realized,which can optimize training samples in the source domain to some extent.Secondly,to further narrow the distance between the source domain and the target domain,the large margin projected transductive support vector machine was applied to complete the transfer of the feature level,thus maximizing the transfer efficiency.Furthermore,the proposed method was applied to the BCI competition dataset (Dataset IIb data set) for offline test and analysis.The results showed that the hybrid transfer learning model achieved significantly better transfer efficiency than the single transfer learning model,and obtained an improvement of the recognition rate with average value of above 70% for different transfer objects.The results also verified the effectiveness and universality of the hybrid transfer learning model.In addition,the online test was carried out based on the established motor-imagery system,which further verified the practicability of the model.
2021 Vol. 42 (5): 629-637 [Abstract] ( 212 ) HTML (1 KB)  PDF (2639 KB)  ( 552 )
638 Research on Interactive Multi-model Set Adaptive Collaborative Filtering Target Tracking Algorithm
KONG De-ming,YANG Dan,WANG Shu-tao
DOI: 10.3969/j.issn.1000-1158.2021.05.15
In order to solve the limitation of the static model set of the traditional interactive multiple model(IMM)algorithm, a interactive multi-model set adaptation collaborative filtering(SAC-IMM)algorithm is proposed. By calculating the model matching probability between the target and different models in the current model set, the best model and the worst model in the current model match are automatically determined. The adaptive process of the model set is achieved by changing the structure of the static model set by using activation, retention and elimination strategies. Compared with the traditional IMM algorithm, the SAC-IMM algorithm proposed has a certain degree of improvement in positioning accuracy. By comparing with other IMM algorithms that have been proposed, the experimented results show that the SAC-IMM algorithm proposed has been optimized for state estimation of speed, acceleration, and turning rate. The proposed method can improve the accuracy of target tracking and positioning to a certain extent.
2021 Vol. 42 (5): 638-644 [Abstract] ( 261 ) HTML (1 KB)  PDF (2224 KB)  ( 593 )
645 Experimental Study on Monochromaticity of 30~160keV Single Energy X-ray Device
WANG Er-yan,JIANG Zheng,GUO Si-ming,WU Jin-jie,YANG Qiang,ZHOU Peng-yue,SONG Rui-qiang
DOI: 10.3969/j.issn.1000-1158.2021.05.16
The single-energy X-ray source is composed of X-ray machine, double crystal Monochromator, standard detector and collimation system. The continuous X-ray produced by X-ray machine is monochromated by Bragg diffraction with double crystal monochromator. The single energy X-ray with energy range of 30~160keV is obtained by adjusting different Bragg angles. In order to study the energy broadening of the calibration device, it is necessary to study the energy resolution of the single energy X-ray produced by the device.The results show that the energy resolution of single energy X-ray produced by Si(220) crystal is 0.91%@30keV and 2.3%@70.6keV, Si(551) crystal is 1.97%@80.1keV and 3.45%@142.6keV. The energy response of lanthanum bromide crystal detector is calibrated and verified by the device. The experimental results show that the device has good energy resolution and can be used in calibration experiments of various types of detectors, measurement of X-ray mass attenuation coefficient and reflectivity measurement of multilayer films.
2021 Vol. 42 (5): 645-649 [Abstract] ( 224 ) HTML (1 KB)  PDF (1015 KB)  ( 219 )
650 Comparison of Methods for Cadmium Analysis in the Rice Matrix Certified Standard Reference Material Based on Uncertainty Evaluation
ZHOU Ming-hui,WU Yan-xiang,CHEN Xi,WANG Song-xue
DOI: 10.3969/j.issn.1000-1158.2021.05.17
Through systematic analysis and evaluation of source, type and contribution of uncertainty for three classical methods including graphite furnace atomic absorption spectrometry (GFAAS), inductively coupled plasma mass spectrometry (ICP-MS) and isotope dilution-inductively coupled plasma mass spectrometry (ID-ICP-MS), it provides basic data for the selection of the best method, and contributed to the process control of the assignment value in the development of cadmium in rice matrix certified standard reference materials. The detected value are 0.478±0.028, 0.488±0.022, 0.485±0.014 mg/kg(k=2) for the GFAAS, ICP-MS and ID-ICP-MS respectively. The ID-ICP-MS proved to be the most reliable method and had the smallest uncertainty for the determination of cadmium in rice, at the same time, the ID-MS method is a benchmark method that can directly trace to the SI unit through the balance weighing and isotope abundance ratio. Therefore, it is an ideal method for determining the cadmium in the rice matrix standard reference materials. Although the uncertainties of GFAAS and the ICP-MS are larger than that of the ID-ICP-MS, the relative expanded uncertainty are only 5.7% and 4.4%, which also could be applied to the value assignment of inter-laboratory studies with different methods.
2021 Vol. 42 (5): 650-657 [Abstract] ( 251 ) HTML (1 KB)  PDF (1447 KB)  ( 130 )
658 Multi-scale Gaussian Kernel Extreme Learning MachineBased on Maximum Correntropy Criterion
LIU Zhao-lun,WU You,WANG Wei-tao,ZHANG Chun-lan,LIU Bin
DOI: 10.3969/j.issn.1000-1158.2021.05.18
In view of the fact that the traditional multi-scale kernel extreme learning machine is sensitive to noise and has a large amount of computation, a multi-scale kernel extreme learning machine which is suitable for Gaussian noise environment is proposed. Firstly, the maximum correntropy criterion is used to replace the traditional minimum mean square error criterion in the multi-scale kernel extreme learning machine to construct the objective function. Secondly, a multi-scale method for randomly generating the scale factors according to the training samples number is applied to the Gaussian kernel function. Finally, the Lagrange multiplier method is used to solve the objective function, and the multi-scale Gaussian kernel extreme learning machine based on the maximum correntropy criterion is derived. Experiments show that the proposed algorithm has higher learning efficiency. Comparing with the traditional multi-scale kernel extreme learning machine, the prediction accuracy on the three UCI benchmark data sets and the application experiment for predicting the f-CaO content of cement clinker,increased by an average of 30.30% and 23.8% respectively.
2021 Vol. 42 (5): 658-667 [Abstract] ( 205 ) HTML (1 KB)  PDF (1720 KB)  ( 180 )
668 Measurement Uncertainty in Flue Gas Pollutants from Thermal Power Plant by Continuous Emission Monitoring System
LIANG Tian-qi,XU Hong,ZHENG Tian-lin,ZHANG Guang-xue
DOI: 10.3969/j.issn.1000-1158.2021.05.19
Continuous Emission Monitoring System (CEMS) has been widely used in the emission monitoring of flue gas pollution of combined heat and power plants, and the data were uploaded to the ecological environment department in real time. In order to evaluate the accuracy and reliability of the CEMS monitoring data of a thermal power plant, uncertainty contributions to the flue gas pollutants measurement results were evaluated, such as zero drift, range drift, indication error, measurement error, standard gas species concentration and measuring repeatability. Results showed that the expanded uncertainties of SO2, NO and NO2 were 2.28mg/m3, 3.50mg/m3 and 0.50mg/m3, respectively, when coverage factor is k=2 and level of confidence is 95%. Because the reference method and CEMS have different sampling methods for flue gas monitoring, the uncertainty component introduced by the measurement error is the largest.
2021 Vol. 42 (5): 668-674 [Abstract] ( 269 ) HTML (1 KB)  PDF (716 KB)  ( 198 )
675 Prediction of SO2 Concentration Based on Mutual Information PSO-LSSVM
JIN Xiu-zhang,LI Jing
DOI: 10.3969/j.issn.1000-1158.2021.05.20
Aiming at the problem of SO2 pollutant emission in thermal power plant, a PSO-LSSVM model prediction method was proposed based on the mutual information.The auxiliary variables of high correlation with the measured inlet concentration of SO2 were selected as the input of the model to realize the prediction of the dominant variable SO2 concentration.The auxiliary variables screened by mutual information had the higher correlation with the auxiliary variables selected by mechanism analysis and Pearson correlation. The auxiliary variables selected by mutual information were used as the input of the LSSVM model and the particle swarm optimization (PSO) was used to determine the parameters of the LSSVM not only reduced the calculation time, but also improved the prediction accuracy. The method was applied to the soft measurement of SO2 concentration in a thermal power plant, and the simulation was carried out by using field data, which showed that the prediction accuracy was higher, the relative error was lower, the prediction trend was closer to the actual value, and the error between the actual value and the predicted value was reduced (the square root error is 2.485 and the average relative error is 0.2603%). It provided the software technical support for the on-site SO2 concentration advance control.
2021 Vol. 42 (5): 675-680 [Abstract] ( 223 ) HTML (1 KB)  PDF (1779 KB)  ( 414 )
681 Research on Model Predictive Control Algorithm of Double-hormone Artificial Pancreas Based on Three-branch Decision Theory
SUN Chao,KONG Xue-hua,GAO Jun,LU Kai-xuan,DU Peng-xiang
DOI: 10.3969/j.issn.1000-1158.2021.05.21
At present, there are few studies on the switching mechanism of double-hormone artificial pancreas subsystems, and existing studies need to be improved in terms of operation efficiency and group control quality. To solve the aboved problem, based on the three decision theory and model predictive control algorithm, a model predictive control algorithm of double-hormone artificial pancreas based on three-branch decision theory was proposed. The algorithm used the three decision theory to design the switching rules among the three subsystems. The rule calculated the decision risk value of switching to each subsystem by designing the cost matrix including economic cost and high and low blood sugar risk, so as to ensure the final switching to the subsystem with the minimum risk value. Finally, 33 virtual patients were experimented on a mature UVA simulation platform and the simulation results were analyzed from various aspects. It was proved that the mentioned algorithm could control blood glucose well within the normal range, and had achieved good results in tracking error, risk index, operation efficiency and group control quality of multiple patients.
2021 Vol. 42 (5): 681-688 [Abstract] ( 222 ) HTML (1 KB)  PDF (1653 KB)  ( 348 )
News
Download
Links
Copyright © Editorial Board of Acta Metrologica Sinica
Supported by:Beijing Magtech