Feature Extraction and the Design of Classifier in Small Set Handwritten Chinese Character Recognition

This paper presents basic concept of feature extraction of handwritten Chinese character, and proposes a new feature extraction named superposition mesh weighting factors strokes extracting algorithm to obtain feature of small set handwritten Chinese character. Basing on the analysis model of RBFNN, an integration RBF classifier is used for small set handwritten Chinese character recognition. Then, the hybrid optimize strategy, which combines the genetic algorithm and the simulated annealing, is adopted to train RBFNN.     

Fast Algorithm of Loads Equalization for Distribution Networks Based on Partition-hierarchical Model

Along with enlargement of scope of distribution, algorithms of reconfiguration for distribution networks become difficult to satisfy real-time require of engineering applications. This paper provides a new algorithm a fast algorithm based on partition-hierarchical model. It is found that partition of loads is relation of the smallest load of vertex that is bigger than average and the biggest load of vertex that is smaller than average. The method reduces the numbers of switches combination by searching smallest load of vertex that is bigger than average and the biggest load of vertex that is smaller than average and realize a fast algorithm of the loads global optimal equalization for distribution networks by transferring the loads. Results of test show that the method presented is applicable and efficient.     

基于人工智能的温室环境控制算法

温室环境控制系统的最大特点是控制对象的精确数学模型很难建立。文章将常规PID控制算法与现代控制理论相结合,对常规PID算法进行改进,得到新型MPT控制算法,并加入模糊控制算法规则,在误差大时,运用模糊算法进行调节,以彻底消除PID饱和积分现象,当误差较小时,采用改进后的PID算法控制输出。同时加入了自适应调节规则和自整定专家系统,建立了基于人工智能的温室环境控制算法。将此算法应用于温室环境控制。系统具有无超调和控制精度高等特点。     

Automatic feature extraction algorithm for porphyrin sensor array images

A feature extraction algorithm for porphyrin sensor array (PSA) images is presented. At the first step, it achieves an automatic tilt correction for PSA images based on radon transform. Then automatic mesh generation is applied to PSA images based on the method of combining projection binarization and mathematical statistics. The areas of porphyrin spots are locating by local threshold segmentation for the images which have achieved mesh generation. Finally, feature values of porphyrin spots are extracted by radius ratio method from the porphyrin spot areas. The algorithm solves the problem that identification of weak signal spots is quite difficult, and the extracted feature values are the maximum point to reflect the information of porphyrin spots. By ammonia, isoprene, aldehydes and other gases tests, the results show that the algorithm can effectively extract feature values of porphyrin spots in PSA images.     

Dynamic optimization design of marine gear transmission system

A dynamic model of transmission system for a marine gear is established considering time-varying mesh stiffness, mesh damping and tooth composite error. The solution efficiency is improved by taking contact line length change of gear pair instead of instantaneous gear mesh stiffness change and making mesh damping and tooth surface equivalent to viscous damping. Multi-objective dynamic optimization design is put into practice to improve the marine gear transmission system using vibration acceleration and mass as target function. The results show that the optimization method can effectively reduce the vibration level and mass of marine gear.     

Dynamic characteristics of multi-stage planetary gear trains in shield machine cutter drive

A purely torsional dynamic model of multi-stage planetary gear trains is developed to investigate the dynamic characteristics of the transmission system in shield machine cutter drive. In the model,phase difference between mesh stiffness and error among each stage caused by different initial mesh positions is considered,and other factors such as component bearing stiffness,time-varying mesh stiffness and error are considered as well. Inherent characteristic analysis shows there are more unique and diverse vibration modes in multi-stage planetary gear trains than in single-stage planetary gear trains. Through dynamic response analysis,time-domain and frequency-domain response of dynamic meshing force of each stage is obtained. As the frequency of exciting force of medium-speed and high-speed stage is in proximity to the natural frequency of the system,harmonic resonance tends to occur,which needs to be paid great attention to. Dynamic factors of each stage are also computed and prepare the ground for the dynamic optimizing design of planetary gear trains.     

The predictive control of an incremental model based on tracking error

An algorithm of incremental model predictive control is proposed to solve a class of industrial production process real-time tracking problems. The basic idea is to establish a predictive model by adopting speed response on the basis of tracking error, and realize the optimization of the second time performance by using the soft constraint of the increment of control quantity and the online rolling optimization method. The theoretical analysis and Matlab simulation results show that the control algorithm can obtain good tracking performance, smooth dynamic response of the controlled object and strong robustness of the closed-loop system.     

MDBPSO算法优化的全波段光谱数据协同冠层SIF监测小麦条锈病

为了从全波段光谱数据中提取对小麦条锈病敏感的特征参量,提高小麦条锈病遥感探测模型的运行效率和精度,本文首先从惯性权重和粒子更新方式两个方面对传统离散粒子群算法(discretebinaryparticleswarmoptimization, DBPSO)进行改进,利用改进离散粒子群算法(modified discrete binary particle swarm optimization, MDBPSO)从全波段光谱数据中优选遥感探测小麦条锈病严重度的特征变量,然后与冠层日光诱导叶绿素荧光(solar-inducedchlorophyllfluorescence,SIF)数据相结合作为自变量分别利用随机森林(randomforest,RF)和后向传播(backpropagation,BP)神经网络算法构建小麦条锈病遥感探测模型,并将其与相关系数(correlationcoefficient,CC)分析法和DBPSO算法提取特征参量构建模型的精度进行对比分析。结果表明:(1) MDBPSO算法比传统DBPSO算法具有更快的收敛速度和更高的寻优精度,改进前后其迭代次数从395次减少到156次,最优适应度函数(optimumfitnessvalue,OFV)值从0.145减小到0.127。(2)采用MDBPSO算法选择特征变量时,RF和BP神经网络两种方法构建的模型精度均高于CC分析法和DBPSO算法,其中RF算法预测病情指数(diseaseindex,DI)值和实测DI值间的检验集决定系数(validation set determination coefficient, R2V)比CC分析法和DBPSO算法分别提高了9%和3%,均方根误差(validation set root mean square error, RMSEV)分别降低了28%和11%, BP神经网络算法预测DI值和实测DI值间的R2V比CC分析法和DBPSO算法分别提高了13%和6%,RMSEV分别降低了21%和10%,利用MDBPSO算法优选特征参量能够提高小麦条锈病的遥感探测精度。(3)在MDBPSO、DBPSO和CC分析法3种特征选择算法中,RF算法构建的模型精度均高于BP神经网络算法,其中RF模型预测DI值和实测DI值间的R2V比BP神经网络算法至少提高了7%,平均提高了9%,RMSEV至少降低了15%,平均降低了20%。以MDBPSO算法优选的特征参量为自变量利用RF方法构建的小麦条锈病遥感探测的MDBPSO-RF模型是小麦条锈病遥感探测适宜模型,该研究结果为进一步实现作物健康状况大面积高精度遥感监测提供了新的思路。     

Generalized conforming quadrilateral membrane element with vertex rotational freedom

A new generalized conforming quadrilateral membrane element is developed by introducing additional rigid vertex rotational freedoms to the normal 4 node quadrilateral membrane element. The rotational stiffness of the present element is not zero so that the singularity of the stiffness matrix is avoided. Numerical analysis of the Cook problem cantilever beam shows that the EAE/Mem element passes the patch test and possesses good convergence, reliability, and precision without a zero energy mode. It provides a lock free solution to the MacNeal beam problem. Numerical results also show that the EAE/Mem element is insensitive to mesh distortion. The results demonstrate that the generalized conforming technique is an efficient tool for developing high performance finite element models.     

Study on Adaptive Sampling and Mesh Generation of Free-form Surfaces Based on Coordinate Measuring Machine

Digitizing of free form surfaces is one of key technologies on the reconstruction of free form surfaces. It is important to know how to sample properly free form surfaces on CMM, which concerns the reconstruction accuracy of free form surfaces and measuring efficiency. The mathematical model of sampling algorithm is proposed based on the concept of the center of gravity. Using the method, the topological information of surfaces can be preserved, and sampling mesh can be self organized. By analyzing and dealing with mesh data received from sampling algorithm, one approach with brevity and higher efficiency is provided on CMM.     

风电齿轮箱传动误差分析及振动噪声预估

针对一级行星两级平行轴风电齿轮传动系统,综合考虑齿轮时变啮合刚度、啮合阻尼、传递误差等因素,建立31个自由度的弯扭轴耦合集中参数动力学模型,采用变步长Runge-Kutta法对系统动力学微分方程进行求解,得出齿轮传动系统各级传动误差;借助软件建立风电齿轮箱刚柔耦合动力学模型,并导入传动误差,采用模态叠加法求得齿轮箱轴承支反力,并将其作为声振耦合模型的边界条件,采用声学有限元法对风电齿轮箱进行振动噪声预估,并与试验结果对比分析,两者吻合良好。     

元动作装配单元误差传递模型及有效路径求解方法

在对机械产品进行"功能-运动-动作"结构化分解的基础上,将影响元动作装配单元装配精度的误差源分为零部件的形状误差、位置误差、装配位置误差和运动误差等四类误差源。引入误差链接模型作为元动作装配单元误差关联关系的基本封装单元,构建结构化误差关联模型——链接网络和链接矩阵,形象描述误差间的耦合嵌套关系。提出基于误差链接模型的装配误差传递路径求解方法,用老鼠迷宫算法搜索所有可能的误差传递路径,以装配精度最高作为判别依据,得到各误差分量的有效传递路径。以蜗杆转动元动作装配单元蜗杆轴线平行度误差有效传递路径为目标,对上述方法进行验证,结果表明该方法能够高效地搜索到所有误差传递路径,并快速获得有效传递路径。该方法的提出为整机装配过程质量预测与控制提供了理论依据。     

Drift error compensation algorithm for fiber optic gyro based on T-S fuzzy modeling

A drift error nonlinear compensation algorithm for Fiber Optic Gyro (FOG) is presented based on T-S fuzzy model with the antecedent parameters identified by G-K clustering algorithm and the error model of T-S fuzzy model with the consequent parameters identified by least square algorithm. The computed results show that this model can compensate the original data effectively, while the error principles of FOG do not need to be understood well. Comparing with the original data, compensation with linear fitting and compensation with neural network, the absolute error of the proposed model reduces by 99%, 96% and 10%, respectively. The error variance reduces by 99%, 98% and 20%, respectively. The results indicate that this proposed algorithm can be simply operated with high precision and easy to realize in engineering.     

Wrinkle modeling based on shape control functions and adaptive subdivision

Realistic wrinkles are an extremely important contribution for enhancing the realism of three dimensional (3D) face model.For different shapes of the wrinkle in three views, this paper proposes three shape control functions for each view, namely, cross-sectional shape control function (CSCF), depth attenuation function (DAF) and width attenuation function (WAF).CSCF is used to realistically and naturally express the furrows and bulges of wrinkles. Depth attenuation function and width attenuation function are applied to control the attenuation variations of the depth and width. Users can create various wrinkles just defining several intuitive parameters. Meanwhile, an image matching technique based on feature is employed to obtain three dimensional wrinkle lines. And a simple adaptive subdivision approach is proposed in the predefined influence region to adjust the resolution around the wrinkle mesh for better results and enhance the efficiency of wrinkle modeling simultaneously. Comparing experiment results with the corresponding image demonstrates that our method can generate realistic and natural wrinkles and can greatly improve the similarity of face model.     

Robust Adaptive Control Based on On-line Identification to Multi-parameters

In system controlled,Sampling these parameters and with parameters for controller in time, it is an object to adaptive control. The article presents a scheme that the structure parameters on it's model and model parameters is robust identified by on line, while to optimal of algorithm on adaptive control is fulfilled. In control scheme, method of intelligent identifying to the structure parameters(levels of model) and algorithm of robust pole placement are introduced also It is advancer for already mentioned scheme. Whole scheme presents new model on adaptive control.     

A Kind of Robust Identification Algorithm Research

A new robust identification algorithm which is based on the estimation of parameter model,i.e.,Optimal Volume Ellipsoid parameter set identification algorithm is presented in this paper.It identifies not only a parameter point(the center of the ellipsoid) but also the error bounds of each parameter.It employs noise bounds in identification algorithm and dynamically modifies coefficients of identification convergent speed and direction in order to attain satisfactory perfomance.Digital simulation results show that is possesses high accuracy,dramatic convergent speed,excellent performance against noise and strong robustness as well as uniform convergent,which is obviously superior to traditional identification schemes.     

Segmentation of Color Image Based on Feature Divergence and Fuzzy Theory

In order to improve general adaptive capability of algorithm,the new color image segmentation algorithm based on feature divergence and fuzzy theory(FDCIS) is proposed.The algorithm introduces feature divergence and fuzzy dissimilarity function into calculation in order to measure the dissimilarity of feature vector,clusters data by means of feature divergence,and accomplishes the merge of image region.The experimental results demonstrate that the color image segmentation result of the proposed approach reduce calculation on large sample of color image,simply and effectively solve over-segmentation of color image,avoid the dependence of the algorithm on initial condition,and hold favorable consistency in terms of human perception.     

State of charge estimation of lithium iron phosphate batteriesbased on adaptive Kalman filters

The Kalman filter algorithm can be used to estimate the state of charge (SOC) of power batteries, however, it easily causes divergence due to uncertain of system noise and its estimation performance is affected by model. An adaptive Kalman filter algorithm is adopted to dynamically estimate SOC of lithium iron phosphate batteries for application in electric vehicles. At first, an equivalent circuit model, appropriate for SOC estimation is built after studying battery models. Then some charging and discharging experiments are carried out for parameter identification and the results are verified. At last, the adaptive Kalman filter algorithm is used on this model for on-line SOC estimation under unknown interfering noise. Simulation results show that adaptive Kalman filter method can correct SOC estimation error caused by tiny model error online, and the estimate accuracy is higher than Kalman filter method. Adaptive Kalman filter algorithm can also correct the initial error. Full-cycle test in electric vehicles proves that the algorithm is appropriate for SOC estimation of lithium iron phosphate battery.     

An improved localization algorithm based on RSSI for landslide monitoring

In order to overcome the existing problems of low automation, high cost and difficult to implement in the area of landslide monitoring, this paper presents a new method of RSSI (Received Signal Strength Indicator)based positioning technology without any one-off instruments to monitor landslide surface displacement. By adding weighted factors, the improved positioning algorithm can estimate the parameters of the path loss model dynamically and calculates the communication distance of network nodes in real-time, which can improve positioning precision and reduce the impact of environmental changes on it. The MATLAB experiments show that, compared with the traditional RSSI based localization algorithm using fixed pass loss model, the improved algorithm could significantly reduce the average error.     

A cooperative localization algorithm for wireless sensor networks

Aiming at the shortcomings of the existing RSS(received signal strength) based localization algorithm for wireless sensor networks (WSN), a cooperative localization algorithm (CLA) is proposed. A reference anchor node is introduced to tolerant some minor error including the node position error. Dixon detection method is applied to remove abnormal RSS values, while the standard deviation threshold of RSS and learning model are introduced to reduce the RSS ranging error and effectively improve the precision. Simulation experiments are performed to evaluate the performance of the proposed algorithm. The results demonstrate that the localization accuracy is improved effectively, while the stability and robustness are better.