Search

To address the issue of model performance degradation in combat intention recognition caused by the long-tailed distribution of battlefield data and the neglect of the spatial dimension information of multivariate time series data, this paper proposes a class balanced spatio-temporal self-attention (CBSTSA) model. By incorporating spatial and temporal attention mechanisms, the model captures interdependencies among features and extracts salient information from both temporal and spatial dimensions. Furthermore, taking the long-tailed distribution of battlefield data into account, a re-weighted class balanced loss function is introduced to train the model. Experimental results show the superiority of our CBSTSA model, e.g. achieving approximately 95.67% accuracy in typical scenarios, surpassing benchmark schemes by 4–5%.

Combining inverse projection and vote accumulation methods, we propose a novel 3-D geometric reconstruction method based on inverse synthetic aperture radar image sequence for triaxially stabilized space targets. Considering that the translation of a triaxially stabilized space target results in a large angular rotation of the radar line-of-sight and an attitude adjustment rotation of the target itself, the projection relationship between the radar imaging plane and a triaxially stabilized space target is constructed first. Subsequently, we perform an inverse projection on the dominant points extracted from each imaging plane to determine the location of the 3-D candidate scatterers and assign a voting value of 1 to each position. By accumulating these vote values, we can obtain the total number of votes for the candidate scatterer position. We treat candidate 3-D scatterers with a large cumulative number of votes as actual scatterers. Thus, the target 3-D geometry can be obtained. Compared with the traditional matrix factorization-based 3-D reconstruction methods, this method is more accurate and ensures that the target pose is consistent with the actual pose. At the same time, compared with 3-D reconstruction methods based on the energy accumulation ideas, this method produces fewer false scatterers and has higher credibility. The experimental results based on the simulated point target and electromagnetic data are presented to validate the effectiveness and robustness of the proposed method.

Relying on the low latency, high reliability, and large bandwidth of 5G networks, the Internet of Things (IoT) can more efficiently promote information interaction and data utilization across remote locations. However, publicly accessible IoT environments involve risks of data and privacy leakage, which threatens the security and feasibility of the cloud-assisted IoT system. Although attribute-based encryption technology provides a fine-grained and secure data sharing solution, it is still challenging to employ dynamic management of multiple groups when device nodes operate cooperatively in dynamic groups. Thus, we propose a dynamic multi-group-supporting policy-hidden secure data sharing scheme in 5G-enabled IoT. Specifically, we exploit a group key management scheme based on the Chinese Remainder Theorem to achieve efficient dynamic group management in IoT and implement concealment of data owners’ access policies to enhance privacy preservation. Moreover, an edge computing layer has been introduced in our 5G-enabled IoT system to entrust decryption to edge servers and cloud servers, thereby mitigating the calculation burden of data users and assisting to fortify resilience against potential collusion attacks. To improve the availability of our system, mechanisms for validating the correctness of the final ciphertext were also incorporated. We further proved the security of the proposed scheme and conducted experimental simulations, which affirm its efficiency and practicability.

With the increasing use of renewable resources and electric vehicles (EVs), the variability and uncertainty in their nature put forward a high requirement for flexibility in AC distribution system incorporating voltage source converter (VSC) based multi-terminal direct current (MTDC) grids. In order to improve the capability of distribution systems to cope with uncertainty, the flexibility enhancement of AC-MTDC distribution systems considering aggregated EVs is studied. Firstly, the charging and discharging model of one EV is proposed considering the users' demand difference and traveling needs. Based on this, a vehicle-to-grid (V2G) control strategy for aggregated EVs to participate in the flexibility promotion of distribution systems is provided. After that, an optimal flexible dispatching method is proposed to improve the flexibility of power systems through cooperation of VSCs, controllable distributed generations (CDGs), aggregated EVs, and energy storage systems (ESSs). Finally, a case study of an AC-MTDC distribution system is carried out. Simulation results show that the proposed dispatching method is capable of effectively enhancing the system flexibility, reducing renewable power curtailment, decreasing load abandonment, and cutting down system cost.

The performance of classifiers plays a crucial role in identifying fault categories under mechanical fault diagnosis task. This paper presents a new intelligent classification approach via NA-MEMD and multiple classifier fusion for fault diagnosis, which comprises four stages. First, NA-MEMD extracts time-domain and frequency-domain features from the original vibration signal. Second, Sensitive feature extraction using improved Fisher’s criterion method. Third, the sensitive feature sets (SFS) are input to Depth Gaussian Restricted Boltzmann Machine (DG-RBM), RNN, and CNN classification methods to attain the complementary benefits and substantial fusion of various classifiers. Final, the Bayesian belief approach is utilized for fusing the classification results of multiple classifiers to obtain the diagnosis results. Experiments on rolling bearing datasets reveal that the presented approach can precisely detect the fault conditions and provide a classification efficiency superior to the single classifiers and ensemble of all classifiers. The experimental results have revealed the efficiency, generalization, and robustness of the multiple classifier fusion.

During the spaceborne inverse synthetic aperture ladar (ISAL) imaging, the radial velocity between the piggyback satellite of ladar and the observed satellite is usually relatively large, which will introduce the inner-pulse Doppler that cannot be ignored in the echo. In this case, the traditional “stop-and-go” assumption will no longer hold, and the compressed result along the range direction would be seriously broaden. Therefore, to obtain a well-focused ISAL image, this article proposed a spaceborne ISAL imaging algorithm for high-speed moving targets based on the inner-pulse Doppler compensation and residue video phase (RVP) correction. In the article, the estimate method of the radial velocity, the inner-pulse Doppler compensation method, and the RVP correction method are given. After inner-pulse Doppler compensation, RVP correction and translational motion compensation, the range-compressed result under the “stop-and-go” assumption can be obtained, and the target can then be imaged by the traditional ISAL imaging algorithm. The simulation results demonstrate the effectiveness of the proposed algorithm.

Deep learning plays an essential role in multidisciplinary research of remote sensing. We will encounter security problems during the data acquisition, processing, and result generation stages. Therefore, secure deep-learning inference services are one of the most important links. Some theoretical progress has been made in cryptographic deep-learning inference, but it lacks a general platform that can be realized in reality. Constantly modifying the corresponding models to approximate the plaintext results reveal the model information to a certain extent. This article proposes a generic post-quantum platform named the PyHENet, which perfectly combines cryptography with plaintext deep learning libraries. Second, we optimize the convolution, activation, and pooling functions and complete the ciphertext operation under floating point numbers for the first time. Moreover, the computation process is accelerated by single instruction multiple data streams and GPU parallel computing. The experimental results show that the PyHENet is closer to the plaintext inference platform than any other cryptographic model and has satisfactory robustness. The optimized PyHENet obtained a better accuracy of 95.05% in the high-resolution NaSC-TG2 database, which the Tiangong-2 space station received.

In the research field of mechanical equipment fault diagnosis, usually only the existing fault types are identified, and the new emerge class of the fault is ignored, however, the new emerge fault class may also occur actually. In order to solve the problem, a novel fault diagnosis model based on deep convolution variational autoencoder network and adaptive label propagation (DCVAN-ALP) is proposed. Firstly, the initial high dimensional features are constructed by using the double tree complex wavelet packet method as the input of the model. Secondly, the convolutional neural network architecture is applied to construct the variational autoencoder, and the local and non-local characteristics of samples are embedded into the loss function for training, which is considered to improve the identification of hidden layer features of the neural network. Finally, t-SNE and the improved label propagation algorithm are adopted to process the hidden features of the neural network, which can achieve the purpose of diagnosing the existing fault class and especially new emerge fault class. Experimental results show that the proposed model can effectively extract the fault characteristics of the vibration signal, and it also has a significantly higher recognition accuracy rate than other typical deep learning methods and traditional classifiers in diagnosing new emerge fault class.

Automatic roof segmentation from airborne light detection and ranging (LiDAR) point cloud data is a key technology for building reconstruction and digital city modeling. In this article, we develop a novel region expansion based L0 gradient minimization algorithm for processing unordered point cloud data, and a two-stage global optimization method consisting of the L0 gradient minimization and graph cut for roof plane segmentation. First, we extract the LiDAR points of buildings from the original point cloud data with a deep learning based method and separate the points of the different buildings using Euclidean clustering to improve the processing efficiency. Second, region expansion based L0 gradient minimization is proposed, which is specially designed for roof plane segmentation from unordered point clouds. To fundamentally avoid the need for empirical parameter tuning in L0 gradient minimization, we propose a multistage coarse-to-fine segmentation process, which further improves the effect of the roof plane segmentation. Finally, graph cut is utilized to solve the jagged boundary and oversegmentation problems existing in the segmented roof planes and produce the segmentation results. We conducted comparative experiments on the Vaihingen and Hangzhou datasets. The experimental results show that the proposed approach significantly outperforms the current state-of-the-art approaches at least 6.7% and 8.9% in roof plane quality index in the Vaihingen and Hangzhou datasets, while showing superior robustness to different kinds of data.

Given a directed graph and a node $v$ , the transitive closure (TC) of $v$ is the set of nodes that $v$ can reach in the graph. TC size is very important in many applications but the cost of TC size computation is high in both time and space, which makes computing accurate TC size not applicable to the scenario where we need to know the TC size quickly for large graphs. Considering that existing approaches are either inefficient or inaccurate, we propose an algorithm, namely buTCS, to efficiently make more accurate estimation. Our approach works in linear time and space. The basic idea is to compute a node’s TC size based on that of its out-neighbors and perform a top-down verification. We further propose two optimizations to improve the estimation accuracy. The experimental results on 13 real datasets show that buTCS achieves better estimation on TC size efficiently.

Semi-polar micro-LEDs have gain increasing interests due to the advantages of polarization control and quantum efficiency improvement. In this work, a novel semi-polar (20-21)-plane micro-LEDs array has been designed and manufactured. In comparison with c-plane micro-LEDs, semi-polar micro-LEDs indicate better electrical and optical performance. The relative EQE of semi-polar micro-LEDs remains at 62% under the injected current density of 775.6 A/cm2, which indicates a reduced efficiency droop due to less polarization in MQWs. It has been further proved by a significant reduction of 55% in emission peak blue-shift under the injected current density from 11.1 A/cm2 to 775.6 A/cm2. In addition, the carrier recombination dynamics and spatial light distribution of semi-polar micro-LEDs with different pixel sizes have been studied. Fast recombination lifetime in smaller size semi-polar micro-LEDs indicates a promising way to be used as a high modulation bandwidth light source. Stable and uniform light distribution in a wider range of spatial azimuths further supports for the semi-polar micro-LEDs as a strong candidate for the applications of high-resolution display and high-speed visible light communication.

Plant identification via leaf images is very meaningful to agricultural information. The existing methods were based on one or two kinds of the three distinct characteristics in leaf images including leaf contours, textures and veins. This limits their recognition performance and scope of application. This paper describes a novel counting-based leaf recognition method, which can directly and effectively combine all of the three kinds of significant characteristics in leaf images. In order to obtain the stable and independent local line responses from leaf contour, texture and vein, elliptical half Gabor is introduced and convoluted with the raw grayscale leaf images, and then maximum gap local line direction patterns are extracted from the local line responses and normalized in direction by cyclically right shifting these patterns until the most numerous bit plane with a value of 1 to the left bit. The histogram of the normalized patterns is calculated and regarded as the counting-based local structure descriptor, and support vector machine is utilized as the classifier. Experimental results on three frequently used leaf databases show that the proposed approach yields a better performance in terms of the classification accuracy, applicability and feasibility in comparison with the state of the art methods.

By combining the transformer feedback and VCO multi-core techniques, this paper presents a 0.5 V low voltage 28.6-to-36.2 GHz dual-core VCO for 5G applications. Based on the transformer feedback topology, with the power-ground interconnect inductor and embedded decouple capacitor, the second harmonic real impedance is achieved at the VCO cross-couple pair source terminal. Accordingly, the common-mode current path is reduced and the signal symmetry improved, suppressing the flicker noise up conversion. With a 4 bit switch capacitors and varactors, the proposed VCO achieves a 28.6-to-36.2 GHz (23.5 %) tuning range, and its flicker noise corner is from 250 to 980 kHz. Fabricated in a 65nm CMOS process, the VCO consumes 4.6mW from a 0.5 V supply voltage with a core area of about $0.24\times 0.5$ mm2. At 31 GHz, the proposed VCO achieves a phase noise of -103 dBc/Hz@1MHz offset, resulting in a figure-of-merit (FoM) of -186.2 dBc/Hz and tuning range figure-of-merit (FoMT) of -193.6 dBc/Hz, respectively.

Autonomous agricultural systems are a promising solution to bridge the gap between labor shortage for agriculture tasks and the continuing needs for increasing productivity in agriculture. Automated mapping and navigation system will be a cornerstone of most autonomous agricultural system. Accordingly, we propose a ground-level mapping and navigating system based on computer vision technology (Mesh Simultaneous Localization and Mapping algorithm, Mesh-SLAM) and Internet of Things (IoT), to generate a 3D farm map on both the edge side and cloud. The innovation of this system includes three layers as sub-systems that are 1) ground-level robot vehicles' layer for conducting frames collection only with a monocular camera, 2) edge node layer for image feature data edge computing and communication, and 3) cloud layer for general management and deep computing. High efficiency and speed of mapping stage are enabled by making the robot vehicles directly stream continuous frames to their corresponding edge node. Then each edge node, that coordinate a certain range of robots, applies a new Mesh-SLAM frame by frame, whose core is reconstructing the features map by a mesh-based algorithm with scalable units and reduce the feature data size by a filtering algorithm. Additionally, the cloud-computing allows comprehensive arrangement and heavily deep computing. The system is scalable to larger-scale fields and more complex environment by taking advantage of dynamically distributing the computation power to edges. Our evaluation indicates that: 1) this Mesh-SLAM algorithm outperforms in mapping and localization precision, accuracy, and yield prediction error (resolution at centimeter); and 2) The scalability and flexibility of the IoT architecture make the system modularized, easy adding/removing new functional modules or IoT sensors. We conclude the trade-off between cost and performance widely augments the feasibility and practical implementation of this system in real farms.

H-polarization, along with E-polarization, indicates the lateral variations of the Earth conductivity, which influence the induced electric field distribution. The coast effect is a typical H-polarization phenomenon that causes local geoelectric field enhancement in coastal areas and significantly affects the geomagnetically induced currents (GIC) distributions in power grids. The influences of H-polarization on geoelectric fields and GIC form the basis for further research on power grid disasters resulting from magnetic storms. In this paper, block and thin-shell models of the coast effect are established, and the electric field distribution in the case of H-polarization is calculated using the finite element method. The results demonstrate the effects of the conductivity, frequency, and distance from the interface of a different conductivity on electric field distortion. Additionally, the relationship between H-polarization and GIC in power grids is investigated, demonstrating that the GIC can be influenced within 100 km in the H-polarization case. The methods and results provide a theoretical basis for GIC risk assessment and development of a control strategy for the power grid.

During a magnetic storm, the induced geoelectric field drives geomagnetically induced currents (GIC) in power transmission networks, railway systems, and pipelines, negatively affecting these systems. In regions with complex geological structures, the lateral earth conductivity changes influence the induced electric field distribution in the earth. H- and E-polarization are two cases of the orientation of the E-field vector relative to the lateral changes. A model of the earth with lateral conductivity changes is established in this paper to examine the effects of conductivity change across a discontinuity on the magnitude of the E-field for the case of the E-field parallel to the discontinuity. The electric field distribution with distance from the discontinuity is calculated, and the relationship between the lateral conductivity changes and electric field distortion is analyzed using the finite element method. In addition, the GIC variation in the power grid due to the lateral conductivity changes is examined. Then, the factors affecting GIC, including conductivity, frequency, and distance, are investigated. The results show that lateral conductivity changes can influence the GIC in power lines running parallel to the discontinuity up to 250 km from the discontinuity. The methods and results are significant for understanding how lateral conductivity changes influence GIC and will improve the accuracy of GIC calculations.

The continuous development in the construction of transportation infrastructure has brought enormous pressure to traffic control. Accurate and detailed traffic flow information is valuable for an effective traffic control strategy. This paper proposes a video-based vehicle counting framework using a three-component process of object detection, object tracking, and trajectory processing to obtain the traffic flow information. First, a dataset for vehicle object detection (VDD) and a standard dataset for verifying the vehicle counting results (VCD) were established. The object detection was then completed by deep learning with VDD. Using this detection, a matching algorithm was designed to perform multi-object tracking in combination with a traditional tracking method. Trajectories of the moving objects were obtained using this approach. Finally, a trajectory counting algorithm based on encoding is proposed. The vehicles were counted according to the vehicle categories and their moving route to obtain detailed traffic flow information. The results demonstrated that the overall accuracy of our method for vehicle counting can reach more than 90%. The running rate of the proposed framework is 20.7 frames/s on the VCD. Therefore, the proposed vehicle counting framework is capable of acquiring reliable traffic flow information, which is likely applicable to intelligent traffic control and dynamic signal timing.

The performance of the pneumatic elements and the micro-controller steadily increases as the price of them decreases. This trend promotes high performance tracking control research on pneumatic servo systems. However, it is very difficult to obtain exact model parameters, which is one of the main obstacles to design a high performance controller. Moreover, in some application cases, the control direction is undetermined because of the possible false or incorrect valve outlets connection. At the same time, the inaccuracy of the proportional valve zero points also degrades the tracking performance of pneumatic servo systems. In this paper, an adaptive backstepping control approach is proposed for a pneumatic position servo system considering unknown model parameters, unknown control direction, and inaccurate valve zero points. The proposed method combines the Nussbaum function and backstepping technique to design a position tracking controller of a pneumatic servo system with the aforementioned uncertainties. By using the Lyapunov method, the designed controller is proved to be stable, and the tracking error asymptotically converges to zero. The experimental results demonstrate the effectiveness and the superiority of the proposed approach as compared with some existing methods, even for negative direction case for which the other methods fail. The performance of the proposed controller is further improved by considering the proportional valve zero points.

Detection of epistatic interactions, which are referred to as nonlinear interactive effects of single nucleotide polymorphisms (SNPs), is increasingly being recognized as an important route in capturing the underlying genetic causes of complex diseases. Its methodological and computational challenges have been well understood, and many methods also have been proposed from different perspectives. Among them ant colony optimization (ACO)-based methods are promising due to their controllable time complexities, heuristic positive feedback search, and high detection power. Nevertheless, there is no comprehensive overview of them so far. This paper, therefore, provides a systematic review of 25 ACO-based epistasis detection methods. First, the generic ACO algorithm, as well as how it is applied to detect epistatic interactions, is briefly described. Then, an in-depth review of ACO-based methods for detecting epistatic interactions is discussed from four aspects, including path selection strategies, pheromone updating rules, fitness functions, and two-stage designs. Finally, this paper analyzes the strengths and limitations of involved methods, provides guidelines for applying them, and gives several views on the future directions of epistasis detection methods.

Neural attention mechanism has achieved many successes in various tasks in natural language processing. However, existing neural attention models based on a densely connected network are loosely related to the attention mechanism found in psychology and neuroscience. Motivated by the finding in neuroscience that human possesses the template-searching attention mechanism, we propose to use convolution operation to simulate attentions and give a mathematical explanation of our neural attention model. We then introduce a new network architecture, which combines a recurrent neural network with our convolution-based attention model and further stacks an attention-based neural model to build a hierarchical sentiment classification model. The experimental results show that our proposed models can capture salient parts of the text to improve the performance of sentiment classification at both the sentence level and the document level.

Compared with the investigation of geomagnetically induced currents (GIC) in power grids, there is less study about the effect of geomagnetic disturbance (GMD) on buried oil and gas pipelines in mid-low latitude areas, which includes GIC and pipe-to-soil potentials (PSP) effects. Therefore, it is of great significance to research the GMD effect in the mid-low latitudes. For this purpose, we performed the observation experiments on GMD influencing the GIC and PSP in China's oil and gas pipelines. The data of GIC and PSP in the pipelines are obtained, which do not pertain to the characteristics of the geomagnetic storms caused by solar activity, as well as the GIC and PSP monitoring data deriving from geomagnetic storm on May 21, 2016. By analyzing the characteristics of waveform, frequency and phase of the geoelectric fields data from Lingyang Geomagnetic Observatory (35°32'N, 118°52'E) 60 km away in northeast from the pipeline observation site, the study corroborated that GIC and PSP variations are driven by tidal geoelectric fields (TGF) during the geomagnetic quiet days in this paper. The mechanism of GIC generated by tidal geoelectric field is analyzed. Likewise, an analysis of GMD data from the Maling Mountain Geomagnetic Observatory (34°42'N, 118°27'E), the GIC and the PSP have been done, it is illustrated that the GIC and PSP effects are caused by TGF and geomagnetic storm induced geoelectric field together during a geomagnetic storm. The results demonstrated that, near large areas of water, the effect of tidal geoelectric field on pipeline corrosion is persistent and serious. TGF impacts the calculation accuracy of GIC in pipelines and PSP. Therefore, we should account for the errors caused by TGF when calculating GIC and PSP in pipelines near seashores and a large body of water.

Ultrasound images of muscle fascicles have been widely used to investigate muscle properties for diagnosis and rehabilitation assessment.The existing automatic fascicle detection and measure methods are based on line detection techniques which do not exactly coincide with the true state of the muscle images. This affects their detection and measure accuracy and depresses their robustness to background interference. In this work, a novel discrete transform namely Belt Linear Summation (BLS) transform is proposed. Unlike the line transform techniques which calculate the sum of pixels on a straight line in images, the BLS transform intends to determine the weighted summation of a belt of pixel values. Based on BLS transform, an automatic fascicle detection and measure method is designed. The performance of the proposed method is compared to the recent automatic fascicle detection and measure methods using both simulated images and clinical images. Experimental results show that the proposed method is robust to background and noise interference, accurate in terms of muscle pennation angle measurement, and feasible for analyzing clinical data.

A network of modular protein complexes inside a cell coordinates many biological processes and is known as protein-protein interaction (PPI) network. APPI network can be modeled as a graph, in which edges represent interactions among proteins, and sub graphs represent protein complexes. Previous methods for protein complex mining from PPI network mainly focused on few topological features like density and degree statistics based on the assumption that proteins inside a complex are highly interactive with each other and thus form dense subgraphs. While this assumption is true for some complexes, it doesn't hold for many others. The important biological information within the protein amino acid sequences, which estimates the interacting property among two proteins for performing a specific biological function is not considered in most of the previous studies. There is a need for algorithms that consider both topological and biological features for correctly identifying protein complexes having varying topological structures and biological patterns inside a PPI network. In this paper, we present an algorithm for detecting protein complexes from interaction graphs. By using graph topological patterns and biological properties as features, we model each complex sub graph by decision tree learners. We use a training set of known complexes to construct decision trees in depth first and BEST FIRST manner using divide and conquer strategy. Splitting criterion, such as information and Gini gain are used in tree expansion process. Training set is divided into subsets and each subset is represented as a branch of tree. Pruning techniques are used to reduce the size of tree. We applied our method to protein interaction data in yeast on two benchmark data sets, i.e., MIPS and CYC2008. According to our results, decision trees achieved a considerable improvement over clique-based algorithms in terms of its ability to recover known complexes by using integrated biological and topological properties.

With outsourcing huge amount of images to cloud, it can reduce the local storage and computational overhead of resource-limited users, but the security concerns still impede the adoption of outsourced image processing services. Fortunately, image processing in the encrypted domain can handle this problem very well and has already attracted considerable attention due to its ability to protect valuable image information from being leaked to untrusted parties. However, the existing schemes incur high computational complexity, and none of them can be used as both global feature and local feature. To this end, we present a method to implement Hahn moments in the encrypted domain by using Somewhat Homomorphic Encryption (SHE) in this paper, named Privacy-Preserving Hahn Moments (PPHM). First, a mathematical framework is proposed to implement the PPHM and image reconstruction in the encrypted domain. Then, the detailed theoretical analysis about data expansion and quantization errors shows that plaintext Hahn moments and plaintext image reconstruction can be implemented by utilizing PPHM over encrypted image. Moreover, security analysis shows that the PPHM can guarantee the image content security. Theoretical analysis and experimental results demonstrate that the PPHM scheme can greatly reduce the computational complexity compared with the discrete cosine transform and the discrete wavelet transform in the encrypted domain. In addition, experimental results show that the PPHM has a better performance in both image reconstruction and image recognition. Furthermore, the suitable value of a scaling factor, which can provide guidance to users, is specified for image reconstruction and image recognition, respectively.

To solve the problem that seriously degraded the kinematics performance may not meet the requirements of on-orbit operation tasks, a coping strategy for a multi-joint multi-type asynchronous failure of a space manipulator is proposed in this paper. Based on the entropy method, a comprehensive kinematics performance index (CKPI) is constructed, which can be regarded as the criterion for solving the artificial joint limits and the optimal locked angle of fault joint. And by solving them, a prevention strategy of serious kinematics performance degradation and a treatment strategy of joint free-swinging failure are established. Through synthesizing the above strategies, a coping strategy for a multi-joint multi-type asynchronous failure is constructed. The contributions of this paper are as follows. The constructed CKPI can make the kinematics performances optimal comprehensively in both operation space and joint space of a space manipulator, the method for calculating the artificial joint limits can solve the coupling problem during calculation, and the constructed coping strategy can ensure that the kinematics performance could meet the requirements of on-orbit operation tasks after multi-joint multi-type asynchronous failure. The correctness and effectiveness of the proposed strategy in this paper are verified by a seven-degree-of-freedom space manipulator.