For the problem of gait stability control for continuous linear walking of a biped robot， a Soft Actor-Critic （SAC） gait control algorithm based on maximum entropy Deep Reinforcement Learning （DRL） was proposed. Firstly， without accurate robot dynamic model built in advance， all parameters were derived from joint angles without additional sensors. Secondly， the cosine similarity method was used to classify experience samples and optimize the experience replay mechanism. Finally， reward functions were designed based on knowledge and experience to enable the biped robot continuously adjust its attitude during the linear walking training process， and the reward functions ensured the robustness of straight walking. The proposed method was compared with other DRL methods such as PPO （Proximal Policy Optimization） and TRPO （Trust Region Policy Optimization） in Roboschool simulation environment. The results show that the proposed method not only achieves fast and stable linear walking of the biped robot， but also has better algorithmic robustness.

Multi-task Joint Learning Model （MJLM） was proposed to solve the performance improvement bottleneck problem caused by the separation of viewpoint-invariant feature and view transformation method in the existing cross-view geo-localization methods. MJLM was made up of a proactive image generative model and a posterior image retrieval model. In the proactive generative model， firstly， Inverse Perspective Mapping （IPM） for coordinate transformation was used to explicitly bridge the spatial domain difference so that the spatial geometric features of the projected image and the real satellite image were approximately the same. Then， the proposed Cross-View Generative Adversarial Network （CVGAN） was used to match and restore the image contents and textures at a fine-grained level implicitly and synthesize smoother and more real satellite images. The posterior retrieval model was composed of Multi-view and Multi-supervision Network （MMNet）， which could perform image retrieval tasks with multi-scale features and multi-supervised learning. Experimental results on Unmanned Aerial Vehicle （UAV） dataset University-1652 show that MJLM achieves the Average Precision （AP） of 89.22% and Recall （R@1） of 87.54%， respectively. Compared with LPN （Local Pattern Network） and MSBA （MultiScale Block Attention）， MJLM has the R@1 improved by 15.29% and 1.07% respectively. It can be seen that MJLM processes the cross-view image synthesis and retrieval tasks together to realize the fusion of view transformation and viewpoint-invariant feature methods in an aggregation， improves the precision and robustness of cross-view geo-localization significantly and verifies the feasibility of the UAV localization.

Aiming at the problem of low accuracy of the existing cross-view image matching algorithms， an Unmanned Aerial Vehicle （UAV） image localization method based on Multi-view and Multi-supervision Network （MMNet） was proposed. Firstly， in the proposed method， satellite perspective and UAV perspective were integrated， global and local features were learnt under a unified network architecture， then classification network was trained and metric tasks were performed in multi-supervision way. Specifically， the Reweighted Regularization Triplet loss （RRT） was mainly used by MMNet to learn global features. In this loss， the reweighting and distance regularization strategies were to solve the problems of imbalance of multi-view samples and structure disorder of the feature space. Simultaneously， in order to pay attention to the context information of the central building in target location， the local features were obtained by MMNet via square ring cutting. After that， the cross entropy loss and RRT were used to perform classification and metric tasks respectively. Finally， the global and local features were aggregated by using a weighted strategy to present target location images. MMNet achieved Recall@1 （R@1） of 83.97% and Average Precision （AP） of 86.96% in UAV localization tasks on the currently popular UAV dataset University-1652. Experimental results show that MMNet significantly improves the accuracy of cross-view image matching， and then enhances the practicability of UAV image localization compared with LCM （cross-view Matching based on Location Classification）， SFPN （Salient Feature Partition Network） and other methods.

Data distributed parallel applications like MapReduce are widely used. Focusing on the issues such as low execution efficiency and high cost of such applications, a case analysis of Hadoop was given. Firstly, based on the analyses of the execution processes of such applications, it was found that the data volume, the numbers of the nodes and tasks were the main factors that affected their execution efficiency. Secondly, the impacts of the factors mentioned above on the execution efficiency of an application were explored. Finally, based on a set of experiments, two important novel rules were derived as follows. Given a specific volume of data, the execution efficiency of a data distributed parallel application could not be improved remarkably only by increasing the number of nodes, but the execution cost would raise on the contrary. However, when the number of tasks was nearly equal to that of the nodes, a higher efficiency and lower cost could be got for such an application. The conclusions are useful for users to optimize their data distributed parallel applications and to estimate the necessary computing resources to be rented in a cloud environment.

Focusing on the higher ratio of processor utilization and lower execution cost of a scientific workflow in cloud, a policy of execution optimization based on task cluster aggregation was proposed. First, the tasks were reasonably replicated and aggregated into several clusters. Therefore, the key tasks could be scheduled as early as possible. Then, the task clusters were aggregated again to facilitate the spare time among the tasks in the task cluster. The experimental results show that the proposed policy can improve the parallelism of workflow tasks, advance the earliest finish time of the whole workflow and it has a significant effect in improving the utilization ratio of processors and lowering the cost of workflow execution.

Semi-Supervised Support Vector Machine using label mean (meanS3VM) for image classification selects a small number of unlabeled instances randomly to train the classifier, and the classification accuracy is low; meanwhile, the parameter's determination always derives much oscillation of the results. In allusion to the above problems, meanS3VM image classification method based on mean shift was proposed. The smoothed image acquired by mean shift was used as original segmented image to reduce diversities of image features; an instance in each smoothed area was randomly selected as unlabeled instance to ensure that it carried useful information for classification and had a more efficient classifier; and the parameters value were also investigated and improved, the grid search method was used for sensitive parameters, the parameter ep was estimated by combining with Support Vector Machine (SVM) mean shift results, so that there will be a better and more stable result. The experimental results indicate that the classification rate of the proposed method to ordinary and noise image can be averagely increased more than 1% and 5%, and it has higher efficiency and avoids the oscillation of the results effectively, which is suitable for image classification.

To improve the running efficiency of business workflow in cloud environment, a policy of replica allocation of cloudy services was proposed. Taking the advantage of social network analysis, the policy specified the central service nodes in a service network based on mining the social network properties such as connectivity and centralization for a service community. The host physical machine of the replica of the central service was specified according to the analysis of logical sequence between the central service and its pre-service, and the usage of other physical machines. The analysis and simulation show that the policy can improve the running efficiency of data intensive business workflow in a cloud environment by averaging the overload of physical machine and reducing the time wasted by long-distance service interaction.

In order to facilitate the execution of the questionnaire survey, and to improve the efficiency of the statistical data collection, this paper proposed a method to realize a questionnaire survey system based on the Android platform running on mobile phone. The survey system was C/S structured. The server included a questions design module based on Java Server Page (JSP), a result statistical module, the database and the C#-based Web service that gave interfaces to access database. The client was implemented on Android platform, which acquired information of the questions from the database, and displayed questions and their options on the screen for users to answer. When the user completed the answer to the questions, the client would write the answers back to the database. This system was first tested on Android emulator, and then on the mobile phone. The testing results indicate that questions survey function has been efficiently realized in the system. Moreover, since the client can run on a mobile device, the survey process can be carried out freely anywhere, anytime, which means that the survey process would not only be convenient and efficient, but also broaden the target clients of the survey system. The proposed system could be adopted by enterprises or organizations to carry out market or social investigations.

To solve the redundancy and high-dimensional problem of the electronic system condition monitoring data, a monitoring data optimization algorithm that combined the sample optimization and features optimization was put forward. Firstly, monitoring data samples were optimized by feature space sample selection algorithm, and the most representative samples were found; then monitoring data characteristics were optimized by KPCA-EDA algorithm after the sample optimization. More recognition information was retained on guarantee that the feature information was enough. Finally, a filter circuit was taken as an example to simulate, and the result shows that this method is effective.

Network Time Protocol (NTP) is a simple, economic and efficient way to accomplish time and frequency synchronization of multiple nodes, while relevant study on the performance evaluation is hard to find in literature, which makes it a question whether to use NTP in application. Concerning this problem, the local network NTP performance and impact of system / network load were measured and analyzed on Windows platform. By comparing time value obtained from IRIG-B time code reader and GetLocalTime Windows API, frequency skew of computer clock signal was approximated. The skew value was close to the value calculated by NTP.

A method was proposed for estimating building heights via registering catadioptric omni-directional image and remote sensing image, which can be applied to large-scale 3D city reconstruction. Firstly, the top edges of building roof were extracted from the catadioptric omni-directional image by using omnidirectional Hough transform. Then the catadioptric omni-directional image and the remote sensing image were registered based on the extracted top edges where the angle consistency nature of horizontal lines in catadioptric omni-directional imaging was used as evidence. Finally, according to the model of catadioptric omnidirectional camera, the building heights were estimated by using the registration results. The proposed method is simple and easy to implement. The experimental results show that the method is effective and the error of estimated building height is fairly small.

In complicated environment, any changes will influence the accuracy of the object detection. Therefore, an algorithm was put forward, which combined the Generalized Gaussian Mixture Model (GGMM) and background subtraction to detect moving objects. The model has a flexibility to perceive environment and model the video background adaptively in the presence of environmental changes (such as radial gradient, background disturbance, shadows and noise). And when it has sudden illumination change，the model can resolve it quickly. In order to meet the realtime requirement, this algorithm adopted the principle, to update every other two frames. The experiments show that it can meet the realtime requirement and detect the moving object accurately.

Patents and journals belong to different knowledge organization systems. To achieve the crossbrowsing and crossretrieval between journal literature and patents，the mapping problem between two classifications Chinese Library Classification (CLC) and International Patent Classification (IPC), must be addressed. According to the survey of the existing methods of classification mapping, this paper discussed a method to achieve the mapping between CLC and IPC based on machine learning. The learner was got by training the corpus identified by the CLC category, with which to classify the corpus identified by the IPC category. The mapping relations can be found after analyzing the classification results. And the comparison experiment proves the effectiveness of this method.

Aiming at the problems of ZigBee tree routing algorithm that the routing may be not optimal and some nodes may use up all the energy because of heavy transmissions, this paper puts forward an improved ZigBee tree routing algorithm based on energy-balanced. The improved algorithm introduces the neighbor table to make sure that the routing is local optimal by considering the routing hops. At the same time, this paper also considers the residual energy of nodes to avoid selecting some nodes with low residual energy in routing selection. The simulation results indicate that this improved algorithm can reduce the energy consumption efficiently, resolve the problem of unbalance load and maximize the lifetime of the whole network.