Search

Aiming at the attitude control problem of fixed wing UAV, this paper introduces S-plane control, which has good control effect in the field of underwater UAV, into the attitude control of UAV. At the same time, aiming at the problem that the coefficient setting of parameters in S-plane control completely depends on experience and cannot be adjusted adaptively, the radial basis function neural network (RBFNN) is introduced, and a neural network S-plane control model which can realize on-line adaptive adjustment of the coefficient of parameters in S-plane control is proposed. The simulation results based on the data of a certain UAV show that compared with the S-plane control, the proposed neural network S-plane control model has the characteristics of fast response speed, strong anti-interference ability, and strong robustness. In addition, it also has the function of adaptive adjustment, which shows good control performance.

In this paper, an implementation method of high sensitivity thin film fiber optic pressure sensor based on air blowing method is proposed. The utility model is characterized in that a film ultra-thin microbubble structure is firstly formed on the surface of the AB epoxy glue solution by using high-pressure air, and then the film portion of the end portion of the microbubble is transferred to a single-mode fiber head with a hollow glass tube at the top. Finally, a thin film structure fiber Fabry-Perot fiber pressure sensor is formed. The experimental results show that the thickness of the thin film layer at the end of the FP sensor is only 16 μm. Correspondingly, the maximum pressure sensitivity can reach 263.15 pm/kPa. Meanwhile, due to the decrease of film thickness, the sensor's resistance to temperature interference is improved. The prepared fiber Fabry-Perot pressure sensor has the advantages of ultra-thin film thickness, high-pressure sensitivity, low cost and good repeatability. Therefore, it has good application value in high sensitivity pressure and sound wave detection.

Mobile edge computing (MEC) is envisioned as a new paradigm by integrating the mobile computing functionality into 5G wireless networks, aiming at empowering communication networks with low-latency services. In general, mobile devices have finite battery lifetime (e.g., machine-type devices) and the energy harvesting is advocated to provide perpetual energy supply for achieving sustainable operation, which is very important for facilitating sustainable computing in future applications. In this paper, we propose a wireless powered MEC network architecture that employs device-to-device (D2D) communications underlaying heterogeneous networks (HetNets) to enable the computational tasks offloading to resource-rich edge servers. A dynamic offloading decision is made to execute the computation tasks. Then, we focus on the energy-efficient offloading scheme, and joint offloading and user association scheme. From the illustrative results, we provide insights for the design of this new network architecture. Furthermore, several open research topics are discussed.

With the higher rate of using cloud storage, protecting data privacy becomes an important issue. The most effective solution is to encrypt data before uploading to the cloud. However, how to efficiently search over data encrypted with different keys is still an open problem. To address this problem, we introduce a new notion of the identity-based encryption with equality test supporting flexible authorization (IBEET-FA). It supports the test of whether two ciphertexts encrypted under the different keys encapsulate the same message, and in the meanwhile supports fine-grained authorization of the test. Based on the equality test on ciphertexts, there is a direct way to support an authorized user to search over ciphertexts of different users, which accelerates secret data sharing among a group of users. Besides, IBEET-FA does not suffer from the complex key management problem of its counterpart in the traditional public key infrastructure. We propose a concrete construction of IBEET-FA and prove it to be securely based on simple mathematical assumptions. The experimental results show that our IBEET-FA scheme is efficient and can satisfy various types of search over encrypted data.

The cyber-physical-social system (CPSS) is a three-layer system framework that combines the human society on the basis of the cyber-physical system (CPS), so that the human society, the cyber world and the physical world are interconnected. In the CPSS, similar profile attributes are matched to socialize and ultimately achieve the purpose of information sharing. However, some personal information may be included in the profile attributes, thus the users' privacy cannot be protected during the process. To meet this challenge, a privacy-preserving profile matching scheme based on private set intersection is proposed in this paper. Multi-tag is utilized to partition the dataset of users to achieve fine-grained profile matching. In addition, the privacy of users is protected by re-encryption technique. Security analysis shows that our scheme is secure against the semi-honest adversary and theoretical analysis of the experiment shows that that the scheme is efficient for profile matching in the CPSS.

Group signature is a cryptography primitive that has been widely researched. It strikes a balance between digital signature and the user's demand for anonymity. A valid member in the group can generate a signature on behalf of the whole group. The public can only know that it was provided by a valid group member and learn nothing about the actual identity of the signer when verifying a group signature. Backes et al pointed out that the existing dynamic group signature schemes implicitly assume that the membership of everyone in the group is open to the public. Thus, they put forward a property called membership privacy for dynamic group signature. In this paper, we design a dynamic group signature scheme with membership privacy on top of Signature Proofs of Knowledge (SPK) and BBS+ signature. Further more, dynamic accumulator mechanism is adopted to revoke a group member's authority to sign. Then, a security analysis demonstrates that the proposed group signature scheme satisfies join-leave privacy. Finally, quantitative analysis and experimental results show that the proposed group signature scheme achieves the fewer signature size and less computation overhead compared with Backes's scheme.

As power consumption of the demand side in regional distribution network shows a diversified development trend, non-intrusive load monitoring and decomposition (NILMD) helps identify the power consumption characteristics. The detailed load information and energy consumption composition offered by NILMD are helpful for dispatching. NILMD can be realized by collecting either transient or steady-state information. However, obtaining transient information requires high acquisition frequency of sensors and it is impractical for users. By collecting steady-state operation information, this paper achieves state extraction and load decomposition for different types of household appliances through affinity propagation (AP), and selected Bayes classification (SBC) clustering methods. Load identification is realized through maximum likelihood estimation. The effectiveness of the proposed methods is verified by a case study.

The World Health organization (WHO) stated that the concept of health has been widened to subjectively experienced dimensions such as fatigue and chronic fatigue syndrome (CFS). With the increasing pressure of the current life, persistent fatigue caused by sustained high-pressure work will not only be hazardous to health, but also give rise to unexpected consequences. In particularly, fatigue driving induced by long time driving has become a leading cause of accidents and death in the transportation. In this study, we investigate electroencephalography(EEG)-based fatigue detection of drivers through the spatial-temporal changes in the relations between EEG channels. EEG signals are firstly partitioned into several segments and the covariance matrices obtained from each segment are fed into a recurrent neural network to extract high-level temporal features. Then, the covariance matrices of whole signals are leveraged to extract spatial characteristics, which will be fused with temporal features to obtain comprehensive spatial-temporal information. Experimental results on a benchmark dataset showed that our method obtained an optimal classification accuracy of 91.042% and outperformed some state-of-the-art methods. These results indicate that our method is reliable and feasible for fatigue detection, which also provides a novel solution for EEG modeling.

Permanent magnet biased axial-radial magnetic bearing (ARMB) can realize axial and radial suspension simultaneously. It has a compact configuration with the volume and weight reduced. However, its magnetic field distribution is quite complicated, its bias and control magnetic field are coupled. In order to design the magnetic bearing accurately, this paper divides the leakage and soft magnetic material routes according to the 2D FE simulation, establishes the accurate magnetic circuit including the leakage and soft magnetic material reluctance. The circuit can be used to observe the flux distribution in each route, verify and regulate the magnetic bearing structure parameter, so the FE computation efforts of the basic structure parameter are significantly reduced. An experimental sample was designed, 3D FE simulation and experiments were processed to test the prototype. The results confirm that the design method is reasonable and correct.

In the design and practical application of highspeed permanent magnet synchronous machine (HSPMSM), there is a multi-parameter, strongly coupled relationship between the machine and the drive. Drive harmonics have a significant impact on machine losses and thus on the overall efficiency. First, a 100kW, 60krpm machine and drive system is modelled accurately to study the strongly coupled electromagnetic field of the motor and drive. The model includes an accurate model of the slot conductor, which facilitates the calculation of AC copper loss. An electromagnetic field-circuit coupling model of the HSPMSM, which enables the study of the high coupling regulation between the machine and the drive. Second, based on the comparison between the ideal sine wave current source and the drive harmonic power supply, the effect of the drive harmonic on various losses of the machine is studied, such as AC copper loss, stator core loss, permanent magnet (PM) and sleeve eddy current loss. The laws and coefficients of the effect of drive harmonics on the various loss of the machine are obtained, which is of great significance for the design and development of HSPMSMs. It provides a useful reference for the design, development and application of HSPMSMs.

Deep Learning (DL) has shown its success and convenience in our daily life, especially for the application in mobile devices. Because of the high quality demand for mobile applications from the consumers, the floating point accuracy of computations especially in deep neural networks (DNNs) has become particularly important. In this paper, we focus on testing the accuracy of floating point in DNN computations in the mobile devices from a statistical point of view. Specifically, two common hypothesis testing methods Z- Test and Student's T- Test are utilized, which cope with different situations. Z- Test is suitable for samples with known population variance and T-Test is suitable for samples without known population variance. Compared to existing scheme that pays more attention on single value comparison in industry, our proposed scheme in field of probability theory is more representative and reaches a more credible results with quantitative measurement, which can be seen from our experiment that based on well-known accelerator HiAI. The most important contribution of this paper is a novel perspective to test floating point accuracy. Future work is to study how to improve the accuracy based on the results.

This paper is about the problem of leader-follower smart ammunition formation (SAF) control. Trajectory time coordination acts an important role on target attack for SAF. With the adoption of a time-coordinate-guidance (TCG) method, the cooperative control model of SAF was built. By comparing with proportional navigation guidance (PNG), simulation results illustrated that a TCG method guided the SAF to attack a fixed target in same time. It proves the possibility of using TCG method for SAF control in real battle environment.

Aiming at the dynamic analysis and formation control of smart ammunition formation, the formation dynamics model and formation control model are established, and the basic dynamic characteristics of smart ammunition formation are analyzed by simulation in this paper. By analyzing the three-dimensional trajectory of leader and follower projectiles, the relative position distance between the leader and follower can be steadily close to the expected value. The simulation results show that the dynamic model and algorithm of smart ammunition formation are ideal. In the future, the time and space synchronization of leader and follower ammunition should be considered, and the cooperative dynamic control technology of smart ammunition formation should be further studied.

The standard system has become one of the important aspects in the modernization of social governance system and governance ability. The establishment of the standard system should not only face the top-level design, but also for the specific implementation. In the process of establishing the standard system by traditional methods, with the help of experts' experience, the future development trend and key development fields are predicted, and the standard system is planned by “brainstorming”. With the continuous progress of information technology, as the “universal language”, standard connects the world by its typical network attributes. With the rapid developed of knowledge graph (KG) technology, establishing the relationship between different standards, and analyzing in depth, we can get the general framework of the standard system. At the same time, combined with the experience of experts and the future development trend of technology, a relatively complete standard system is formed. The results show that the method adopted in this paper has certain guiding significance for the establishment of a scientific standard system.

This paper copes with the problems of l 1 -gain analysis and control for switched positive systems with a dwell time constraint. State-dependent switching adopted in this paper satisfies minimal dwell time constraint to avoid possible arbitrarily fast switching. By constructing multiple linear co-positive Lyapunov functions, sufficient conditions of stability and l 1 -gain property are developed under the proposed switching strategy. Then, an effective state feedback controller is designed to ensure that the closed-loop system is positive and has l 1 -gain property. Finally, a simulation example is given to illustrate the effectiveness and potential of the proposed method.

Deep neural networks (DNNs) have been proved to be popular and successful in our daily life. For mobile devices with the services from DNN, it is needed to equip a DNN accelerator to employ fast and accurate DNN models. Different DNN accelerators may have different accuracy performance, which directly affects the user experience. To measure the accuracy of the DNN accelerators, there only exists indirect approach by comparing to the third-party platform with 32-bit float format. In this paper, metamorphic testing is employed to measure the accuracy of the DNN accelerator for the first time. Specifically, our scheme utilizes the mathematical properties of the operators in deep neural networks to measure the accuracy of the DNN accelerator, getting rid of the dependence on the third-party platform. Besides, we design an objective metric to quantitatively measure the accuracy performance. It can be used not only to facilitate the direct competition between different accelerators but also to detect defects in the implementation of the DNN accelerator. An experiment on two well-known DNN accelerators, HiAI and Snapdragon Neural Processing Engine (SNPE), is conducted, and the results show that the metamorphic distance of the former is smaller than that of the latter in terms of their accuracy performance with 16-bit float format on typical Ops, such as softmax and convolution.

Electroencephalography (EEG)-based depression detection has become a hot topic in the development of biomedical engineering. However, the complexity and nonstationarity of EEG signals are two biggest obstacles to this application. In addition, the generalization of detection algorithms may be degraded owing to the influences brought by individual differences. In view of the correlation between EEG signals and individual demographics, such as gender, age, etc., and influences of these demographic factors on the incidence of depression, it would be better to incorporate demographic factors during EEG modeling and depression detection. In this work, we constructed an one-dimensional Convolutional Neural Network (1-D CNN) to obtain more effective features of EEG signals, then integrated gender and age factors into the 1-D CNN via an attention mechanism, which could prompt our 1-D CNN to explore complex correlations between EEG signals and demographic factors, and generate more effective high-level representations ultimately for the detection of depression. Experimental results on 170 (81 depressed patients and 89 normal controls) subjects showed that the proposed method is superior to the unitary 1-D CNN without gender and age factors and two other ways of incorporating demographics. This work also indicates that organic mixture of EEG signals and demographic factors is promising for the detection of depression.Clinical relevance-This work indicates that organically mixture of EEG signals and demographic factors is promising for the detection of depression.

With the gradual maturity of the domestic financial quantitative trading market, more and more investors have recognized the quantitative investment because of its relatively stable income. Investment trading is not interfered by human beings. Analyzing the historical data of market trading has the characteristics of high probability of winning. This paper will take s300etf(159912) as an example, combine the double moving average and Counter-market operating indicator(CDP), explain the application law of the indicator, and discuss how to better grasp the market and achieve higher return on investment through the two indicators. Stock data is a classical financial time series, so Long Short-Term Memory(LSTM), as a classical model in the cyclic neural network model, can predict the trading index through LSTM. In this paper, the backtrader platform is used to simulate the transaction of historical data, and the benefits of the simulated transaction are discussed.

Identity-based cryptosystems mean that public keys can be directly derived from user identifiers, such as telephone numbers, email addresses, and social insurance number, and so on. So they can simplify key management procedures of certificate-based public key infrastructures and can be used to realize authentication in blockchain. Linearly homomorphic signature schemes allow to perform linear computations on authenticated data. And the correctness of the computation can be publicly verified. Although a series of homomorphic signature schemes have been designed recently, there are few homomorphic signature schemes designed in identity-based cryptography. In this paper, we construct a new ID-based linear homomorphic signature scheme, which avoids the shortcomings of the use of public-key certificates. The scheme is proved secure against existential forgery on adaptively chosen message and ID attack under the random oracle model. The ID-based linearly homomorphic signature schemes can be applied in e-business and cloud computing. Finally, we show how to apply it to realize authentication in blockchain.

Wireless sensor networks (WSNs) distribute hundreds to thousands of inexpensive micro-sensor nodes in their regions, and these nodes are important parts of Internet of Things (IoT). In WSN-assisted IoT, the nodes are resource constrained in many ways, such as storage resources, computing resources, energy resources, and so on. Robust routing protocols are required to maintain a long network lifetime and achieve higher energy utilization. In this paper, we propose a new energy-efficient centroid-based routing protocol (EECRP) for WSN-assisted IoT to improve the performance of the network. The proposed EECRP includes three key parts: a new distributed cluster formation technique that enables the self-organization of local nodes, a new series of algorithms for adapting clusters and rotating the cluster head based on the centroid position to evenly distribute the energy load among all sensor nodes, and a new mechanism to reduce the energy consumption for long-distance communications. In particular, the residual energy of nodes is considered in EECRP for calculating the centroid′s position. Our simulation results indicate that EECRP performs better than LEACH, LEACH-C, and GEEC. In addition, EECRP is suitable for networks that require a long lifetime and whose base station (BS) is located in the network.