This PDF file contains the front matter associated with SPIE Proceedings Volume 11929, including the Title Page, Copyright information and Table of Contents

Centered on parallel factor analysis(PARAFAC) and support vector machine(SVM), a method of centrifugal pump(CP) fault diagnosis is proposed. A non-stationary multi-fault mode single-source signal feature extraction method is proposed, based on the theory of parallel factor analysis, i.e. spatial information is introduced in time-frequency analysis, and after parallel factor analysis, the spatial signal, the spectral signal (mode2) and the time-domain signal (mode3) are generated. The latter two can obviously describe the natural or fault state of the mechanical equipment by comparing the experimental analysis, and this function is used to evaluate the mapping relationship between the centrifugal pumps in different states and their respective mode2 and mode3. A total of six components in mode2 and mode3 are used as feature vectors, which are input to the SVM classifier, optimized by the improved particle swarm algorithm for fault classification, by constructing an experimental platform to collect the nonlinear multi-fault mode feature signals of centrifugal pumps. Compared with the wavelet packet energy features, the proposed method is more convenient for feature extraction, the classification accuracy of the proposed classifier is significantly improved, and the complexity of the proposed classifier is not significantly increased.

In this research, multi-agent networked system cooperation is considered based on the leader-follower model, and application to smart irrigation has been carried out. Leader-follower model is proposed by the consideration of IoT sensors at the irrigation area. For the watering areas, we adopt leader-follower networked system to save water resource and electrical power driving water pump. Selected area or electrical pump from all irrigation area, it has the role of leader over the neighborhood. In order to provide conservative water control, highest water level is chosen as the leader. Leader-Follower modelling has been derived and verified. In order to calculate the reward by applying Leader-Follower model, Shapley value is considered and compared with the conventional result. From the simulation result, we verified good performance by using Leader-Follower model. The obtained result can be linked with Internet of Things (IoT), it makes us to manage and control water with web application as well.

Compared with the traditional task allocation strategy, the proposed model considers the dynamic, limited and diversity resources in the process of offloading policy selection and analyzes the key factors in task offloading scheduling. Thus, the concept of opportunistic computing capacity is introduced in the opportunistic network, and the specific mechanism of task offloading is proposed. Considering the implementation method, the task execution time is studied in detail, and the mathematical model and solving algorithm are given. Finally, the simulation results show that the proposed method has more advantages than the traditional task scheduling method in terms of offloading efficiency and resource utility ratio.

In recent years, smart grids have developed rapidly. As an important terminal of the smart grid, the reliability and safety of intelligent electric meters play an important role in the development of smart grids. However, in practical applications, intelligent electric meters have problems such as inconsistent software architecture, less penetration testing involved in software testing, and difficulty in finding vulnerabilities. In order to solve the above-mentioned problems, this paper analyzes the software architecture of intelligent electric meters and establishes a hierarchical system model. Then, actively analyzes the weaknesses, defects or vulnerabilities in the software to detect the design vulnerabilities and software defects in the intelligent electric meter, gives qualitative or quantitative test and verification results for its fault tolerance and ease of recovery, and establishes a complete software reliability and safety testing system, and develops a prototype system, which has good guiding significance and implementation value for the software reliability and safety testing of intelligent electric meters.

With the development of optical imaging technology, when a hyperspectral camera continuously obtains hyperspectral imageries in a short period of time, temporal information will also become available. Because the spatial-spectral information and historical trajectory play different roles in moving target detection, fusion of different information has become a good framework for moving target detection. However, it is difficult to choose a suitable scale to represent different types of moving targets which have different scales and different moving speeds. To solve this problem, multiscale spatial-spectral detection and multi-scale temporal variance filter is proposed. Experimental results demonstrated that adopting multi-scale fusion strategies in spatial-spectral map and historical trajectory map (MSSH) in real hyperspectral moving target datasets can effectively improve the detection performance of different moving targets.

Image stitching adopts image registration and image fusion technology, which combines several images with overlapping areas into a wide-view, high-resolution panoramic image. Image Stitching is widely used in remote sensing image processing, medical diagnosis, military detection and many other fields. When the image is stitched, the feature points extracted by methods based on grayscale, frequency domain and image features are too redundant, and the feature points are not comprehensive enough. In order to solve the above problems, this paper proposes a network framework for extracting feature points of image stitching based on deep learning, which is called SA-LIFT. First, the image is divided into similar and non-similar areas by extracting model. Second, an improved LIFT model is proposed, which uses three convolutional networks to extract image feature points. The experimental results show that the method used in this paper has good effects in improving comprehensiveness and reducing redundancy when extracting the feature points of image stitching.

The microwave source queue of high-power microwave heating system has only binary detection of whether there are running or not. Online real-time detection of the working status and changes of the microwave source queue is necessary in order to improve the economy and reliability of its queue. A fault detection method of microwave source queue in high power microwave heating system based on wavelet analysis is presented. Firstly, the fault of microwave source queue is judged according to the change of electric field intensity. Secondly, by acquiring the continuous wavelet transform of the input signal and output signal at multiple scales to calculate the residual signal generated by the extreme value, and the detection is completed according to the change of the residual signal. Finally, the wavelet time-frequency spectrum of the residual signal is drawn to obtain the feature information and perform visual analysis of the detection results. Three fault states of microwave source are simulated respectively, the effectiveness of the method is verified by simulation and result analysis. The continuous wavelet transform analysis method is used to detect the fault of the microwave source queue, which can accurately detect the fault location and time of microwave source queue online.

Through the way of device atomic abstraction and dynamic cross configuration control, a smart hotel RCU (room control unit) system with re-configurable control logic is realized and the principle is described in detail; which solves the problem that the former RCU with fixed control logic needs to be redesigned to meet the diversified intelligent scene design requirements of different hotel. At present, this method has been successfully deployed in many smart hotels such as GreenTree Inns Chains, Xian Hotel and Tupai Hotel. Also the method can be extended to be used in other IOT application fields to realize the dynamic management and logic reconstruction of IOT nodes.

To effectively remove the complex noise in the pulsed eddy current signals of downhole double-layer pipe, we propose a monitoring signal denoising algorithm (MSDA) of eddy current for outer casing pipe of double-layer pipe. First, we analyze the noise component of the eddy current detection signal, which comes from the electromagnetic detection instrument MTD-J that detect the double-layer pipe of a high sulfur gas field. Then we establish the noise model to describe the noise and design a deep neural network model to obtain the noise model parameters. Finally, we use the noise model to denoising the noise of the eddy current detection signal. Experimental results show that MSDA improves the SNR. It is better than Wavelet denoising (WD) and Empirical Mode denoising (EMD).

There is a certain lag in the conventional means of measurement and supervision, periodic system inspection, equipment maintenance, etc., so the on-line verification of equipment cannot be achieved. At the same time, the voltage-withstand test is a key link of the project verification, and its results indicate the insulation level and safety performance of the tested samples. The abnormality in device will cause loss of control of the test results, missed inspections, false inspections, etc., and damage to the tested samples and the equipment itself. In this paper, a method for on-line monitoring and fault positioning of abnormality in multi-epitope voltage-withstand test device is proposed. By collecting the leakage current in the same batch of meter voltage-withstand test and comparing the consistency and difference of the characteristic distribution of leakage current among different epitopes, the running state of the device can be monitored online, fault positioning and abnormal diagnosis can be realized. the method proposed in this paper is verified by actual test data.