Search

Abstract: Today''s industrial control equipments on site become more automated, complex and continuous; in order to ensure normal, reliable and efficient operations of those equipments, and to ensure the safety of site personnel, condition monitoring and fault diagnosis must be carried out on those on-site equipments. In this paper, based on the studies of traditional fault diagnosis theories and modern fault diagnosis theories, it uses fault tree analysis (FTA) to establish the fault diagnosis model of tread line control system. [Copyright &y& Elsevier]

Abstract: A mathematical model of optimal scheduling scheme for gas pipeline network is established, which takes maximum benefits of pipeline network service provider as objective function after comprehensively considered the resources of gas field, downstream gas utilizing conditions, operating parameters of compressor stations and work conditions of pipeline system etc. Evolutionary-based particle swarm algorithm has been investigated, and a new hybrid particle swarm algorithm with embedded chaotic search has been presented to the optimization solution of pipeline network scheduling. Improved algorithm makes use of the ergodicity of chaos to improve the capability of precise search and keep the balance between the global search and the local search, which is able to avoid the premature convergence problem effectively. This paper took a gas network with 17 nodes, 16 pipelines, 2 compressor stations as example, and make the decision on the optimal scheduling scheme for gas pipeline network. The analysis of example demonstrates the effectiveness and applicability of the established model and algorithm. [Copyright &y& Elsevier]

Abstract: The blind signal separation which has many potential usage, especially in the respects of such as wireless communication, medical analysis, geographical exploration, and image enhancement and recognition is a developing basic research direction. This paper mainly studies the application of it in speech processing and proposes a mathematical description of the blind signal separation problem in including mathematical model for blind separation problem and basic assumptions to ensure the blind separation to be solvable. It focuses on the analysis of the blind signal separation based on instantaneous mixed mode. It also proposes speech blind signal separation algorithm based on instantaneous mixed circumstance and performs computer emulation and signal-to –noise-ration analysis to confirm the effectiveness of the algorithm. [Copyright &y& Elsevier]

Abstract: This paper investigates the dynamics of closing-day futures prices on crude oil traded in the NYMEX and IPE markets. Empirical results indicate that volatility of oil futures is characterized by long memory. We compare models with normal and Student-t distributions and find that FIGARCH model seems to fit the data better. Based on the criteria of mean absolute forecast error (MAE), mean squared error (RMSE) and the hit rate (HR), our data analysis shows that support vector machine (SVM) outperforms the competing models in most situations of volatility forecasting, which confirms the theoretical advantage of SVM. [Copyright &y& Elsevier]

Abstract: In order to achieve the predicted error and stability in the optimal process, by using BP neural network model, aiming at the improvement of heredity and variation in evolutionary computation, and shortcomings in evolutionary learning algorithm, this paper carries out the evolutionary training on the BP network and obtains good results. [Copyright &y& Elsevier]

Abstract: Human action in video sequences can be seen as multidimensional arrays, i.e., tensors, and action video of gray-scale can be directly modeled as a typical 3rd-order tensor. In this paper, we exploit an efficient appearance-based method for recognition of action video volumes using silhouettes based on high order SVD. We use the silhouette ensembles to form a 3rd-order tensor for each action video, search a tensor-based principal component analysis (PCA) to do dimensionality reduction and analyze action volumes. Through experiments on a common database, we see our method not only preserves the spatiotemporal information of the video, but also outperforms the 1-mode PCA in appearance-based recognition of human actions. [Copyright &y& Elsevier]

Abstract: Recent researches have well established that sparse signal models have led to outstanding performances in signal, image and video processing tasks. This success is mainly due to the fact that natural signals such as images admit sparse representations of some redundant basis, also called dictionary. This paper focuses on learning discriminative dictionaries instead of reconstructive ones. It has been shown that discriminative dictionaries, which are composed of sparse reconstruction and class discrimination terms, outperform reconstructive ones for image classification tasks. Experimental results in image classification tasks using examples from the Caltech 101 Object Categories show that the proposed method is efficient and can achieve a higher recognition rate than reconstructive methods. [Copyright &y& Elsevier]

Abstract: In this paper, a novel method, KFP-SVM, of generating synthetic samples with kernel model is proposed to deal with imbalanced data sets. In this method, the majority samples are selected by removing those points which invade the minority samples, and the minority samples are mapped into a high-dimension space with a kernel function which is called kernel space. Then the distance of the minority samples in kernel space is calculated. The nearest neighbors of every minority samples are found out and linear combination is used to generate synthetic samples. Finally these synthetic samples are mapped inverse into input space, so that the primary synthetic samples and a SVM model will be obtained. A group of experiments in UCI data sets have shown that this novel method with kernel model to generate the synthetic samples could improve the performance of the SVM under the criterion of F-Measure and AUC, compared with SVM,SMOTE-SVM,DEC-SVM. [Copyright &y& Elsevier]

Abstract: The available cases with class labels are not enough for building costumer credit rating model in practice, especially for the newly established system in which old customers’ data do not exist. For fast and correctly evaluating costumer credit rate with limited labeled data sets under parallel working environment, a method for building classifier based on non-myopic active learning is proposed in this paper. Based on an information-theoretic criterion, our method selects a set of samples at each iteration, and the criterion function is proved to be submodular, which guarantees to find the near-optimal samples for labeling by experts. The experimental results on telecom costumers’ credit data sets show the model built by the non-myopic active learning method can reach the same accuracy as passive learning when the amount of the initially labeled data is small. This can significantly reduce the annotation cost for credit risk evaluation. [Copyright &y& Elsevier]

Abstract: Recognizing Chinese person names has long been a hot topic in the field of natural language process. Past researches and experiments find that some certain sorts of names like two-character person names cannot be well recognized. This paper presents a hierarchical method for automatically recognizing Chinese person names from Chinese language material. This hierarchical method outperforms than previous method at recognizing those names, since it integrates information provided both by decision tree method and conditional random field (CRF) model soundly. Experimental results show that our method is effective by achieving 95.2% precision and 92.7% recall rate. [Copyright &y& Elsevier]

Abstract: To achieve inter-domains access authorization, role mapping relationships between different domains need to be established in multi-domain collaborative system based on RBAC. This paper introduces access control model achieving secure interoperation in multi-domains, through expanding and reconstructing access control policy. Policy and algorithm are proposed, which separated roles generated in the original model are released, including Static Release and Dynamic Release, in order that the original model can be applied in multi-domain access control system based on RBAC without time constraint. [Copyright &y& Elsevier]

Abstract: Dynamic Voltage Restorer (DVR) based on the line voltage compensating is proposed in this paper. A single-input and single-output model of DVR is established, with the load voltage as control object,the duty ratio of the power devices as control variable, and the utility voltage and load current as interferences. Also the transfer functions from variale to control object and from interferences to control object are derived. Transfer functions of different system control strategies are derived and conclusion is drawn that the system steady error is caused by two interferences. A compound controller of both utility voltage and load current feed forward plus load voltage feed back is thus proposed, which guarantees the fast dynamic response and system stability, and eliminates the system steady error. Theoretical analysis is verified through simulations and tests. [Copyright &y& Elsevier]

Abstract: This paper investigates dynamic handling and stability of coach. Based on the analysis and simplification of coach structure and load, and by the application of ergonomics principle and vehicle system dynamics, the models of coach components, such as tire, suspension, passenger body, and so on, are respectively established, and ultimately the coach dynamics models of poly-degree of freedom are founded. In addition, experiments with SX6120 are carrying out on a certain road. Through comparing the simulation data with those of tests, the results show that the coach dynamics model is correct. Using the model built, the responses of coach handling and stability are simulated under high-speed avoiding obstacle conditions. The simulating results indicate that handling and stability of coach is obviously improved when lateral stabilizer rod is fixed. And the dynamics performances of the coach with different stiffness lateral stabilizer rod are different. [Copyright &y& Elsevier]

Abstract: Construction project involves multitudinous stakeholders, how to balance the interests and reduce the conflicts among these stakeholders to constitute effectively harmonious construction team has became a critical factor to reduce construction resources’ waste. With the use of game theory, this paper created disintegrating disharmonious construction alliance model and integrating harmonious construction alliance on the basis of former work. It concluded conditions and priorities of the two Unions and provided policy Suggestions. [Copyright &y& Elsevier]

Abstract: The laneway cost determination of mine construction engineering is the key point of the coal mine construction, its rapidity and the accuracy have obvious effect in construction stage, it is advantageous in reducing time and saving management cost in this stage. For this aim, this paper establishes a cost estimation model of mine construction by neural network firstly, then takes the cost estimation of laneway project as an example, selects suitable project characteristics, Uses MATLAB provided the BP neural network model to predict the laneway construction estimated cost. The result indicates that the neural network method can achieve the goal better and more quickly. [Copyright &y& Elsevier]

Abstract: With the elementary and secondary schools as its research objects, this paper introduces the conception of parameters, establishes the parameterized information models of the plot and the building as well as the corresponding relation between them by setting up the appraisal values and the priority of the environmental impacts and the building functions, and implements computer-aided architecture design on this basis. It realized human-computer interaction and has valuable significance for architects and the planning department. [Copyright &y& Elsevier]

Abstract: Facilities management is an integrated management model which surpasses property management and ensures the building environment features. Based on the basic theory of facilities management and 2010 Shanghai World Expo project management practice, this paper analysis the characteristics of its facilities management operations, establishes the facility breakdown structure and the three organizational system from the point of views of the scope definition and organization designing, in order to establish a benchmark for the operation and management of similar projects organizational planning. [Copyright &y& Elsevier]

Abstract: This paper analyses the application of a switching volatility model to forecast the distribution of returns and to estimate the Value-at-Risk(VaR). We calculate the VaR value for 10 SSE stocks. the VaR values calculated using the switching regime beta model are preferred to other methods. The Proportion of Failure tests confirm this result. [Copyright &y& Elsevier]

Abstract: This paper discusses a batch arrival queue with delayed single working vacation. Whenever the system becomes empty at a normal service completion instant, the server will enters into the delay period and will return into a normal busy period immediately without vacation period if an arrival occurs during such a period. Otherwise he will go for a single working vacation. We derive the queue size distribution at random epochs and departure epochs. We also derive some important performance measures of this model. Finally, several numerical examples are presented. [Copyright &y& Elsevier]

Abstract: As a modeling tool, Petri net has been widely used. However graphic Petri nets can give the visual description, but not simulate execution process and automatically generate reachability graph. Maria is an effective and comprehensive analyzer for P/T Nets. It can complete dynamic analysis of Petri nets. In this paper, the structure and relationship of graphic Petri nets and Maria modeling language are discussed in detail. On this basis, a conversion algorithm from graphic Petri nets to description with Maria modeling language is presented. Finally, the functional modules of graphic Petri nets analyzer based on Maria are designed and introduced. This analyzer offers a convenient way for analyzing a Petri net model by Maria. [Copyright &y& Elsevier]

Abstract: No a detection method can detect all of intrusion behaviors, a perfect IDS should be a unity of multiple detection means. This paper''s object is to establish an intrusion detection system model based on data mining technology, which can detect known intrusion effectively and has a good capacity to recognize unknown data pattern which can’t be detected effectively in traditional IDS. The experiments proved that the method is effective to detect intrusion such as scanning and DoS. [Copyright &y& Elsevier]

Abstract: In this paper, by the use of similarity function and the Hadamard product, we proposed a new algorithm to solve the consistency problem in group decision making with unbalanced linguistic information. Compared to the previous studies of Herrera, the characteristic of the model is the briefness of the calculation process and the expert opinion is adjusted in accordance with the concise principle. [Copyright &y& Elsevier]

Abstract: The paper presents a new intrusion detection model combining misuse detection and anomaly detection mode, and makes a research into the key technology of the model based on data mining theory. In the model, the association rules data mining algorithm is applied to establish abnormal behavior rule set for misuse detection to detect known intrusion rapidly. And the minimum dissimilarity clustering analysis algorithm is used to establish normal behavior rule set for anomaly detection to detect new unknown intrusion. Research of the model of intrusion detection based on data mining is made on KDD99 dataset. The experiment shows that the new model can improve true positives, decrease false positives and detect new intrusions. [Copyright &y& Elsevier]

Abstract: The paper analyzed the present situation of modern distance education and then described the necessity that carried out distance education personalized services firstly. It constructed a personalized modern distance education model based on web data mining and described the basic design idea and functions in detail. For the distance education organization, this model will provide a certain theoretical basic and reference value to build a personalized modern distance education system. [Copyright &y& Elsevier]

Abstract: Most of the current detection models for software behavior focus on the integrity detecting of its running path, but they rarely consider detecting the function realization integrity in each stage of the software running path. Therefore, dynamic credibility detection model based on scene mining for software behavior called SCDM is proposed in this paper. First of all, this model describes the integrity of various stages’ function implementation by introducing the concept of scene. Secondly, the model used association rules and sequence pattern algorithm in scene mining. Both of the algorithms can extract the hidden and valuable association information from the captured scene information set. And the mining result is used for credibility detection of software behavior. Finally, the results of simulation experiments indicate that CDMBSM model has a good detection performance. [Copyright &y& Elsevier]

Abstract: In dynamic Grid virtual organizations, domains involved must agree upon cross-domain access control policies for their shared resources. Currently, almost all access control mechanisms in Grid security can not support collaborative privilege management. In this paper, we propose a negotiation-based cross-domain access control mechanism, which can be divided into two stages: multi-domain negotiation stage to solve permission assignments to domains and bidomain negotiation stage to make concrete cross-domain access control policies based on user attributes between two domains. Negotiation models and processes are presented as well as examples to illustrate their application. [Copyright &y& Elsevier]

Abstract: Fruit trees growth models are helpful to understand and predict fruit trees growth processes and their reactions to the environment, and models also can be used for the resources utilization analysis, and for providing support for production decision. This paper introduced the basic procedures for establishing growth model and the application of common models in fruit trees growth processes, summarized the main problems existing in the research and gave future expectation, in order to provide reference for the fruit trees cultivation and management. [Copyright &y& Elsevier]

Abstract: In many broadcast encryption environment it is important to protect both confidentiality of encrypted messages and the identity information of authorized users. But most of current identity-based broadcast encryption (IBBE) schemes fail to protect the privacy of their users. In this paper, we propose the formal definition for privacy-preserving IBBE (PPIBBE). Then, we give a new PPIBBE scheme that is adaptively secure in the standard model. [Copyright &y& Elsevier]

Abstract: The terrestrial laser scanning system can quickly acquire precise and dense point cloud from object surface. But in actual application, such as: visualization, data simplification, usually needs to constitute object mesh model used point cloud. This paper studied the construction method of mesh model based on point cloud. Considering the characteristics of one station terrestrial laser scanning point cloud, the improved triangulation growth algorithm model was presented. Firstly one station point cloud was projected onto a sphere with the method of central projection, constructed triangulation model on the sphere to ensure the correct adjacency relation between the points; then according to four limiting conditions of space Delaunay, realized mesh model construction for the bronze statue of Chairman MAO in Fuxin. [Copyright &y& Elsevier]

Abstract: The mathematical model of the quantum reversible circuit is unitary matrix. Matrix can better reflect the quantum state evolution and the physical properties of quantum computation. Matrix transformation based algorithm for 4-qubit reversible logic circuit synthesis is proposed in this paper. The algorithm skillfully takes advantage of the matrix representation of quantum circuit and the circuit transformation rules of the adjacent matrix to construct 4-qubit circuit. Compared with the constructive algorithm which is popular at present, the number of the gates in circuit constructed by our algorithm is less. [Copyright &y& Elsevier]

Abstract: In this paper, a swarm of agents is created to solve clustering ensemble problem. The agent''s individual model is a key point, thus the complex global goal can be obtained through agent pursuing its private goal. Finally the swarm of agents is applied to 4 collections of diversity clusterings from different clustering algorithms. The consensus clustering has higher similarity to the “true clustering” than initial clusterings. [Copyright &y& Elsevier]

Abstract: How to build a quick and cost-effective energy supply network which is as easy as the station is a new complex planning problem. Firstly the paper discusses the problems that are related to the construction of charging facilities. Then the analytic hierarchy process (AHP) is applied to establish the evaluation models for the functional area and the service window. Finally, the entire energy supply network gets a reasonable and orderly control, with good social benefits achieved. [Copyright &y& Elsevier]

Abstract: The mainly purpose of this paper is to simulate the charging and discharging process of internal melt ice-on-coil thermal storage system based on the exited experiment. Its mathematic model is built and the results of the simulation and experiment are compared for verifying the simulation. The further analysis about the influence of whole ice storage and melt process is made from thermal resistance, diameter and pipe material. Analysis indicates that making diameter larger can make heat transfer better. Larger the heat transfer coefficient of the pipe can improve thermal efficiency, but the heat resistance will not be reduced obviously as the coefficient increase. [Copyright &y& Elsevier]

Abstract: Aiming at distributed energy system (DES) with energy storage device, the mathematical models of major equipments are established in this paper. With the objective functions of the daily minimum operation cost and the minimum equipment quantity, a mixed integer nonlinear multistage objective function is established, which is solved by the means of dynamic programming and penalty function and then the system optimization is completed. According to the case of a hospital in Tianjin, the optimized result is compared with the other two modes, which are DES without energy storage device and sub-production system (the existing system of the hospital). The result shows that the system with energy storage device is better than the system without energy storage device, and the economy of DES is superior to sub-production system significantly. [Copyright &y& Elsevier]

Abstract: A theoretical analysis of novel direct expansion heat pump system incorporating a Photovoltaic/Thermal (PV/T) solar collector acting as the evaporator is presented in this paper. Based on energy balance of each parts of the vacuum-tube-PV sandwich and the four main components of the heat pump system, mathematical models are developed to evaluate the energy performance of the heat pump and the electrical and thermal performances of the PV cells, under the climatic conditions of the City of Nottingham, UK. The simulation results indicate that the average monthly thermal and electrical efficiencies are 0.752 and 0.155 respectively, and the average monthly COP reaches 5.35 [Copyright &y& Elsevier]

Abstract: The development of industry economy cans not get away from nature environment to prop up. It is big, researching on economic growth and environment pollution. Economic growth action has been object for many countries. The traditional mode of economic growth led to serious environmental pollution. Facing three major scientific issues that called economic growth, resources and environment, we will strengthen marginal analysis of economic growth on theory. According to traditional marginal analysis model, we propose a model of marginal analysis model of economic growth and environmental pollution. Analysis by adding environmental pollution, this paper, we propose a cost of economy growth, and gives control of economic growth for the cost of environment pollution. [Copyright &y& Elsevier]

Abstract: Recent studies on mathematics model of activated sludge were reviewed in this paper. The current applied research of urban wastewater treatment process model and development of software were mainly introduced. And the prospect and development direction are also introduced. [Copyright &y& Elsevier]

Abstract: Based on the study of enterprise middle-level management performance appraisal index, this paper studies the fuzzy mathematics’ application in the middle-level management performance evaluation, sets up the fuzzy comprehensive evaluation model of middle-level management performance and carries out the empirical study, which offers theoretic and realistic basis for strengthening middle-level management performance appraisal. [Copyright &y& Elsevier]

Abstract: With China''s rapid economic growth, low-carbon economy is an inevitable choice in China. Development of lowcarbon service sector is not only the objective requirement of sustainable economic development in the 21st century, but also an inevitable choice for service industry development in Hunan. This paper try to explore under the new service development model transformation goals and the financial crisis after the new situation of China''s service sector, firstly introduced the status and characteristics of the low-carbon service sector in Hunan, and then analysis problems and put forward countermeasures. [Copyright &y& Elsevier]

Abstract: There are many factors involved in slope stability evaluation and the principle ones acts differently under different slope environmental conditions. So the evaluation of slope stability demands for consideration for all kinds of factors generally with clear understanding of primary and secondary factors. Based on the theory of ant algorithms and cellular automaton, this paper puts forward a cellular ant algorithm and mathematic description with MATLAB to optimize the accuracy of slope stability evaluation and verifies the applicability of the algorithm through real project and achieves good results. [Copyright &y& Elsevier]

Abstract: The predictable evaluation of the risk of roadbed geological hazards depends largely on the reliability of the fundamental information and the rationality of the mathematical model. As all the factors of the geological environment at different regions are different in the quality, the same index is likely to attach to the different fatalness grade at the different regions. Therefore, the application of the classical mathematical model to dividing the uniform standards for measuring still belongs to one of the challenging frontiers in the current international earth science field. In this paper we attempt to establish the mathematical model of contribution-ratio predication method in the predictable evaluation of the risk of roadbed geological hazards to determine the critical rainfall in which geological hazards take place, hoping to provide the theory reference for the predication and warning problems of the risk of roadbed geological hazards. [Copyright &y& Elsevier]

Abstract: The stability evaluation of the surrounding rock of tunnel with small clear spacing is an important step in a tunnel construction process. In this paper an indicator system was proposed to evaluate the stability state of the central rock pillar between twin tunnel with small clear spacings; then an AHP (Analytic Hierarchy Process) model was constructed to calculate the weights of indicators; and finally, with fuzzy mathematic method, a fuzzy comprehensive evaluation model was set up to evaluate the stability state of the central rock pillar between twin tunnel with small clear spacings. Project cases showed that this method is practical and effective. [Copyright &y& Elsevier]

Abstract: In the paper, least squares support vector machine trained by chaos particle swarm optimization (CPSO-LSSVM) is applied in the classification. Chaos particle swarm optimization algorithm is introduced to select the suitable training parameters of least squares support vector machine. That is because chaos particle swarm optimization algorithm can avoid being trapped in local optimum and gaining the global optimum quickly in the process of search by combining particle swarm optimization with chaos queues. The experimental results demonstrate that least squares support vector machine trained by chaos particle swarm optimization is best classification results in the three classification algorithms. [Copyright &y& Elsevier]

Abstract: This paper introduces the concept of finite automata, formal definitions and basic composition, finally described finite automatáapplication in the ATM by a example. [Copyright &y& Elsevier]

Abstract: In order to identify the damage of architectural structure and ensure the quality of architectural structure, least squares support vector regression is proposed to identify the damage of architectural structure in the paper. The experimental data are given to study the evaluation performance of the proposed method. In the experimental data, the inherent frequencies are used as the features and the deep of crack is used as evaluation index. The experimental results show that damage identification results of least squares support vector regression are better than those of support vector regression. [Copyright &y& Elsevier]

Abstract: Activated sludge model (ASM) is applied into represent complex biochemical reaction between various contaminants in wastewater and microbial wastewater treatment system. Based on activated sludge ASM 1 model in this paper, the aerobic process COD is separated into dissolved easily biodegradable organics, soluble inert organics, particulate non-biodegradable organics, particulate slow-biodegradable organics and particulate inert matter produced by heterotrophic biological decay. With a volume constant and completely mixed batch reactor, the process of simplified ASM 1 model was described. The simplified model improves the speed because of the less parameter with a similar simulation result used by ASM 1 model. But it is very necessary to adjust the model parameter in different industrial wastewater treatment . [Copyright &y& Elsevier]

Abstract: Based on the summarization of research status quo of water resources carrying capacity, the paper establishes the model for calculating water resources carrying capacity and takes the “carrying population” as its integration index. The model is tested and calculated by taking Baoding as practical example. [Copyright &y& Elsevier]

Abstract: Activated sludge plays a very important position in biological wastewater treatment process, mathematical models and computer technology have been applied in activated sludge process widely. After fitting the experimental data and developing microbial growth kinetics model, the identification model based on biological wastewater treatment process characteristics became hot topic. This changed activated sludge process design into research of the dynamics activated sludge process for the target of high efficiency of low-power. Then, corresponding activated sludge process commercial software was developed. The first activated sludge process software based on ASM1 model, SSSP (simulation of single sludge process), was developed in 1988. With the modeling advanced development, various activated sludge process commercial software were developed, such as EFOR, STOAT, SASS PRo, DSP and GPS-X. In this paper the activated sludge software are introduced and some advises are proposed. [Copyright &y& Elsevier]

Abstract: Many studies on tourist carrying capacity (TCC) at present are concentrating on ecological protection. The determination of TCC should consider the supply capacity of basic facilities in resort as well as ecosystems, such as transport vehicles, hotels and restaurants. In this paper we will introduce an evaluating index which is called “level of service” (LOS) to represent the utilization of basic facilities, and develop a function of LOS and the number of tourists. Calculation model of TTC is derived corresponding to critical value of LOS. Then from the view of ecological protection, the other TTC model is developed according to the size of tourism area and maximum tourist density parameters. Optimal TTC equals the minor of the values calculated from the two TTC models. Lastly, YunTai resort, a World Geology Park in HeNan province is used as example to explain the model. Results show that carrying capacity of LOS is 8511 and smaller than ecological carrying capacity of 11730. Measures should be taken to promote the LOS for tourists. [Copyright &y& Elsevier]

Abstract: This paper presents a moment invariants based three-dimensional face recognition method under sparse representation framework. An improved moment invariants descriptor is constructed and the discriminative ability in face recognition is validated on NPU3D and FRAV database. A series of partial faces are obtained by segmenting each face model according to the geodesic distance from nose middle point. Proposed moments invariants is computed and stacked as the feature of each face model. Finally, sparse representation is used in the classification step. The result on GavabDB shows that this method is very promising compared with ICP baseline. [Copyright &y& Elsevier]