Search

This article provides a targetless extrinsic calibration method between camera and low-resolution 3D LiDAR. We carried out real-world experiments on a multi-sensor platform containing a Velodyne VLP-16 LiDAR and a camera. Ground-truth extrinsic matrix is generated by a state-of-the-art target-based method. Regarding the calibration error, the rotation and translation RMSE of our method are 0.59° and 3cm respectively. Codes are avaible at https://github.com/gitouni/Targetless-LiDAR-camera-calibration. NB: This article has been accepted for publication in IEEE Sensors Journal. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. Citation Information: 10.1109/JSEN.2023.3263833

Background: Little is known about the long-term risks of stroke and ischemic heart disease (IHD) in women who had a hysterectomy alone (HA) or with bilateral oophorectomy (HBO) for benign diseases, particularly in China where the burden of cardiovascular diseases (CVD) is high. We assessed mean levels of cardiovascular risk factors and relative risks of stroke and IHD in Chinese women who had a HA or HBO. Methods: A total of 302 510 women, aged 30 to 79 years were enrolled in the China Kadoorie Biobank from 2004 to 2008 and followed up for a mean of 9.8 years. The analysis involved premenopausal women without prior cardiovascular disease or cancer at enrollment. We calculated adjusted hazard ratios for incident cases of CVD and their pathological types (ischemic stroke, hemorrhagic stroke, and IHD) after HA and HBO. Analyses were stratified by age and region and adjusted for levels of education, household income, smoking status, alcohol consumption, physical activity, body mass index, systolic blood pressure, diabetes, self-reported health, and number of pregnancies. Results: Among 282 722 eligible women, 8478 had HA, and 1360 had HBO. Women who had HA had 9% higher risk of CVD after HA (hazard ratio, 1.09 [95% CI, 1.06–1.12]) and 19% higher risk of CVD after HBO (1.19 [95% CI, 1.12–1.26]) compared with women who did not. Both HA and HBO were associated with higher risks of ischemic stroke and IHD but not with hemorrhagic stroke. The relative risks of CVD associated with HA and HBO were more extreme at younger age of surgery. Conclusions: Women who had either HA or HBO have higher risks of ischemic stroke and IHD, and these risks should be evaluated when discussing these interventions. Additional screening for risk factors for CVD should be considered in women following HA and HBO operations, especially if such operations are performed at younger age.

This paper proposes a proportional-integral (PI) event-triggered control strategy for networked nonlinear systems subject to unmatched uncertainties. The influences of the unmatched uncertainties and external disturbances on the output channel are attenuated by an extended-state-observer-based output feedback controller through a state transformation. The outputs of the controller are transmitted in an event-triggered fashion by comparing the differences between the current sampling error and its integration with respect to time. We prove the stability of the closed-loop event-triggered control system by regulating the trajectories of the observation error and the tracking error into a compact set. Further, the relationships between control performance and designed parameters are investigated quantitatively. In addition, the nonexistence of Zeno behavior is guaranteed by proving the existence of a nonzero minimum inter-event time. Finally, the effectiveness of the proposed PI-type event-triggered control strategy in reducing signal transmission rate is validated through a numerical example and various experiments.

This research provides a novel approach for detecting multi-legged robot actuator faults. The most significant concept is to design the Fault Diagnosis Generative Adversarial Network (FD-GAN) to fully adapt to the fault diagnosis problem with insufficient data. We found that it is difficult for methods based on classification and prediction to learn failure patterns without enough data. A straightforward solution is to use massive amounts of normal data to drive the diagnostic model. We introduce frequency-domain information and fuse multi-sensor data to increase the features and expand the difference between normal data and fault data. A GAN-based framework is designed to calculate the probability that the enhanced data belongs to the normal category. It uses a generator network as a feature extractor, and uses a discriminator network as a fault probability evaluator, which creates a new use of GAN in the field of fault diagnosis. Among the many learning strategies of GAN, we find that a key point that can distinguish the two types of data is to use the hidden layer noise with appropriate discrimination as the excitation. We also design a fault location method based on binary search, which greatly improves the search efficiency and engineering value of the entire method. We have conducted a lot of experiments to prove the diagnostic effectiveness of our architecture in various road conditions and working modes. We compared FD-GAN with popular diagnostic methods. The results show that our method has the highest accuracy and recall rate.

Purpose This paper aims to focus on lane detection of unmanned mobile robots. For the mobile robot, it is undesirable to spend lots of time detecting the lane. So quickly detecting the lane in a complex environment such as poor illumination and shadows becomes a challenge. Design/methodology/approach A new learning framework based on an integration of extreme learning machine (ELM) and an inception structure named multiscale ELM is proposed, making full use of the advantages that ELM has faster convergence and convolutional neural network could extract local features in different scales. The proposed architecture is divided into two main components: self-taught feature extraction by ELM with the convolution layer and bottom-up information classification based on the feature constraint. To overcome the disadvantages of poor performance under complex conditions such as shadows and illumination, this paper mainly solves four problems: local features learning: replaced the fully connected layer, the convolutional layer is used to extract local features; feature extraction in different scales: the integration of ELM and inception structure improves the parameters learning speed, but it also achieves spatial interactivity in different scales; and the validity of the training database: a method how to find a training data set is proposed. Findings Experimental results on various data sets reveal that the proposed algorithm effectively improves performance under complex conditions. In the actual environment, experimental results tested by the robot platform named BIT-NAZA show that the proposed algorithm achieves better performance and reliability. Originality/value This research can provide a theoretical and engineering basis for lane detection on unmanned robots.

Accurate path tracking and stability are the main challenges of lateral motion control in mobile robots, especially under the situation with complex road conditions. The interaction force between robots and the external environment may cause interference, which should be considered to guarantee its path tracking performance in dynamic and uncertain environments. In this article, a flexible lateral control scheme is considered for the developed wheel-legged robot, which consists of a cubature Kalman algorithm to evaluate the centroid slip angle and the yaw rate. Furthermore, a fuzzy compensation and preview angle-enhanced sliding model controller to improve the tracking accuracy and robustness. Finally, some simulations and experimental demonstrations using the four-wheel-legged robot (BIT-NAZA) are carried out to illustrate the effectiveness and robustness, and the proposed method has achieved satisfactory results in high-precision trajectory tracking and stability control of the mobile robot.

Parallel electrical manipulator systems have become prevalent in attitude simulation in recent years to achieve a high precision attitude tracking. However, the attitude tracking performance is affected by internal uncertainty, external disturbances, and uncoordinated performance of the electric actuators. In this article, a coordinated attitude control strategy that combined the distributed active disturbance rejection control (ADRC) framework with a consensus algorithm is proposed for the parallel electrical manipulators. We indicate that the proposed coordinated attitude control strategy solves the uncoordinated problem of the electric actuators, and compensates the disturbances that include internal uncertainty and external disturbances. Meanwhile, a closed-loop control is achieved with a guaranteed stable tracking performance. The control strategy is applied to a 3-DOF parallel electrical manipulator system, and comparative experimental results indicate that the proposed coordinated ADRC consensus control strategy can achieve a better tracking performance than the strategy with only ADRC.

This article presents the results of a comparative study of personal ornaments from Xiaogushan Cave in the interregional and regional context of the formation of modern behavior. Xiaogushan is a Paleolithic and Neolithic site in Northeast China. In the Upper Paleolithic layers of the site, apart from tools, personal ornaments were found— pendants made from animal teeth, and a decorated bone disc. The date of the site is a matter of debate; ornaments from layers 2 and 3 date to ~30 ka BP. Like other bone artifacts (harpoon, needles, point), and together with types of stone tools and lithic technology, they mirror the local process of Middle to Upper Paleolithic transition. We focus on similarities between the Xiaogushan ornaments and Upper Paleolithic pendants from northern China and Eurasia in general, attesting to modern behavior during the transitional period and being an important marker of the spread of Upper Paleolithic innovations from the centers to the periphery. Xiaogushan is the fi rst Upper Paleolithic industry in Northeast China known to date, and demonstrates skills and symbolic behavior typical of the initial Upper Paleolithic. The Xiaogushan pendants follow the general tendencies, while being specifi c markers of the evolution of symbolic behavior in Eastern Eurasia.

Fine-Grained Visual Categorization (FGVC) is a challenging task due to the large intra-subcategory and small inter-subcategory variances. Recent studies tackle this task through a weakly supervised manner without using the part annotation from the experts. Of those, methods based on bilinear pooling are one of the main categories for computing the interaction between deep features and have shown high effectiveness. However, these methods mainly focus on the correlation within one specific layer but largely ignore the high interactions between multiple layers. In this study, we argue that considering the high interaction between the features from multiple layers can help to learn more distinguishing fine-grained features. To this end, we propose a High-Order-Interaction (HOI) method for FGVC. In our HOI, an efficient cross-layer trilinear pooling is introduced to calculate the third-order interaction between three different layers. Third-order interactions of different combinations are then fused to form the final representation. HOI can produce more discriminative representations and be readily integrated with the two popular techniques, attention mechanism and triplet loss, to obtain superposed improvement. Extensive experiments conducted on four FGVC datasets show the great superiority of our method over bilinear-based methods and demonstrate that the proposed method achieves the state of the art.

Purpose This paper aims to focus on the environmental perception of unmanned platform under complex street scenes. Unmanned platform has a strict requirement both on accuracy and inference speed. So how to make a trade-off between accuracy and inference speed during the extraction of environmental information becomes a challenge. Design/methodology/approach In this paper, a novel multi-scale depth-wise residual (MDR) module is proposed. This module makes full use of depth-wise separable convolution, dilated convolution and 1-dimensional (1-D) convolution, which is able to extract local information and contextual information jointly while keeping this module small-scale and shallow. Then, based on MDR module, a novel network named multi-scale depth-wise residual network (MDRNet) is designed for fast semantic segmentation. This network could extract multi-scale information and maintain feature maps with high spatial resolution to mitigate the existence of objects at multiple scales. Findings Experiments on Camvid data set and Cityscapes data set reveal that the proposed MDRNet produces competitive results both in terms of computational time and accuracy during inference. Specially, the authors got 67.47 and 68.7% Mean Intersection over Union (MIoU) on Camvid data set and Cityscapes data set, respectively, with only 0.84 million parameters and quicker speed on a single GTX 1070Ti card. Originality/value This research can provide the theoretical and engineering basis for environmental perception on the unmanned platform. In addition, it provides environmental information to support the subsequent works.

Cats infected with feline calicivirus (FCV) often display oral ulcers and inflammation of the upper respiratory tract, which can lead to death in severe cases. Antiviral therapy is one of the most effective ways to control FCV infection. Natural compounds in Chinese herbal medicines and medicinal plants provide abundant resources for research on antiviral drugs. In this study, we found that icariin (ICA), formononetin (FMN) and caffeic acid phenethyl ester (CPAE) show low cytotoxicity towards F81 cells, that the three natural compounds have apparent antiviral effects on FCV in vitro, and that they can inhibit different FCV strains. Then, we found that ICA and FMN mainly function in the early stage of FCV infection, while CAPE can function in both the early and late stages of FCV infection. Finally, we found that ICA has an antagonistic effect on FMN and CAPE in FCV infection, and FMN has a synergistic effect with CAPE against FCV infection. Our results showed that ICA, FMN and CAPE may be potential drug candidates for FCV-induced diseases.

Purpose When it comes to the high accuracy autonomous motion of the mobile robot, it is challenging to effectively control the robot to follow the desired trajectory and transport the payload simultaneously, especially for the cloud robot system. In this paper, a flexible trajectory tracking control scheme is developed via iterative learning control to manage a distributed cloud robot (BIT-6NAZA) under the payload delivery scenarios. Design/methodology/approach Considering the relationship of six-wheeled independent steering in the BIT-6NAZA robot, an iterative learning controller is implemented for reliable trajectory tracking with the payload transportation. Meanwhile, the stability analysis of the system ensures the effective convergence of the algorithm. Findings Finally, to evaluate the developed method, some demonstrations, including the different motion models and tracking control, are presented both in simulation and experiment. It can achieve flexible tracking performance of the designed composite algorithm. Originality/value This paper provides a feasible method for the trajectory tracking control in the cloud robot system and simultaneously promotes the robot application in practical engineering.

PurposeRenal fibrosis is a pathological state in the progression of chronic kidney disease. Early detection and treatment are vital to prolong the survival of patients. Renal puncture examination represents the gold standard examination method for renal fibrosis, but it has several limitations. This study aims to evaluate the diagnostic performance of a novel PET radiotracer, [68Ga]Ga-fibroblast activation protein inhibitor (FAPI)-04, which specifically images fibroblast activation protein (FAP) expression for renal fibrosis.MethodsAll patients underwent renal puncture before receiving [68Ga]Ga-FAPI-04 PET/CT imaging. They then underwent [68Ga]Ga-FAPI-04 PET/CT and immunochemistry examinations, and the data obtained were analyzed.ResultsThe [68Ga]Ga-FAPI-04 PET/CT examination results showed that almost all patients (12/13) exhibited increased radiotracer uptake. The maximum standardized uptake value (SUVmax) in patients with mild, moderate, and severe fibrosis was 3.92±1.50, 5.98±1.6, and 7.67±2.23, respectively.Conclusion Compared with renal puncture examination, non-invasive imaging of FAP expression through [68Ga]Ga-FAPI-04 PET/CT can quickly show the patient’s bilateral kidney condition with high sensitivity. This can facilitate the evaluation of the patient’s disease progression, diagnosis, and the development of a treatment plan.

Two ionic liquid (IL) thermomorphic systems were designed. The thermoregulated property exhibited by the IL systems was utilized for the recovery of Au(III) from acidic medium. The cloud point order of the two ILs was proposed and evidenced via a combination of experimental measurements and quantum chemical calculations. The impacts of salt ions and acid on the thermomorphic phase behavior of the IL–H₂O binary mixture system were investigated. After comprehensive comparison, a systematic study of Au(III) extraction by the [Suc][Tf₂N]₂ (succinyl bis(trifluoromethylsulfonyl)imide) system was carried out because of its outstanding hydrophobicity and extraction ability. Then, a comparison between traditional liquid–liquid extraction (TLLE) and homogeneous liquid–liquid extraction (HLLE) was made during the process of optimization of various extraction parameters. In addition, the anion-exchange mechanism in the extraction process was revealed. The thermoregulated [Suc][Tf₂N]₂ system presented an excellent selectivity for Au(III) in a solution of coexisting multimetallic ions. Finally, Au(III) was almost completely stripped from the loaded IL phase by oxalic acid. The extraction efficiency of Au(III) using regenerated IL still maintains a surprising stability after five turns of cycling experiments. Hence, the dicationic IL [Suc][Tf₂N]₂ was regarded as a potential extractant for the recovery of gold.

This paper provides a legged stable walking control strategy based on multi-sensor information feedback about BIT-NAZA-II, a large load parallel hexapod wheel-legged robot developing for the problem of vertical contact impact and horizontal sliding of heavy leg robot in complex terrain environments. The BIT-NAZA-II robot has six legs and six wheels, and the wheels are installed on the foot-end. The wheels of each foot-end for the legs of the robot are locked when walking with legs. In order to realize the smooth transition between swing phase and stance phase, the leg motion is divided into different stages for control by state machine switching controller based on event (SMSCE). In the Z-direction, in order to avoid the shaking of the body caused by the contact impact at the moment of contact between the foot-end and the ground during the walking of the robot, an active compliance controller (ACC) based on impedance control (IC) is applied to solve the problem of contact impact. Moreover, in the X-direction, the swing leg retraction (SLR) based on Bezier curve (BC) is introduced to generate the foot-end trajectory of the robot, which solves the slip problem of the heavy leg robot and improves the horizontal stability. Finally, the control strategy of stable walking is respectively verified by the simulations and experiments. The results show that the ACC based on IC can effectively reduce the contact impact between the foot-end and the ground in the Z-direction and improve the stability of body. Besides, the anti-sliding ability is realized after introducing SLR based on BC in the X-direction, and we also verify that stable walking control strategy is effective, which provides a reference value for the stable walking of heavy leg robot in complex terrain.

AIM We aimed to use data-driven glucose pattern analysis to unveil the correlation between the metrics reflecting glucose fluctuation and beta-cell function, and to identify the possible role of this metric in diabetes classification. MATERIALS AND METHODS In total, 78 participants with type 1 diabetes and 59 with type 2 diabetes were enrolled in this study. All participants wore a flash glucose monitoring system, and glucose data were collected. A detrended fluctuation function (DFF) was utilized to extract glucose fluctuation information from flash glucose monitoring data and a DFF-based glucose fluctuation metric was proposed. RESULTS For the entire study population, a significant negative correlation between the DFF-based glucose fluctuation metric and fasting C-peptide was observed (r = -0.667; P

Reinforcement learning (RL) depends on carefully engineering environment rewards. However, rewards from environments are extremely sparse for many RL tasks, challenging for the agent to learn skills and interact with the environment. One solution to this problem is to create intrinsic rewards for agents and to make rewards dense and more suitable for learning. Recent algorithms, such as curiosity-driven exploration, usually estimate the novelty of the next state through the prediction error of dynamics models. However, these methods are typically limited by the capacity of their dynamics models. In this paper, a random curiosity-driven model using deep reinforcement learning is proposed, which uses a target network with fixed weights to maintain the stability of dynamics models and create more suitable intrinsic rewards. We integrate the parametric exploration method for further promoting sufficient exploration. Besides, a deeper and more closely connected network is utilized for encoding the pixel images for policy-gradient. By comparing our method against the previous approaches in several environments, the experiments show that our method achieves state-of-the-art performance on most but not all of the Atari games.

Background High drug prices can limit treatment access. Patented drugs can be sold for prices far higher than the costs of production. This analysis aimed to determine prices currently feasible to treat COVID-19, HIV, and Hepatitis B or C, with WHO/FDA/NICE approved oral medications assuming competitive generic manufacture. Methods Data on Active Pharmaceutical Ingredients (API) exported from India were collected from the Panjiva database ( www.panjiva.com ) – an online database of global exports – and were used to calculate current weighted mean cost/kg of API. Target prices were calculated based on the per-pill cost of API, plus costs of manufacture ($0.01/pill), 10% profit margin, and assumed 27% tax on profit. We selected a range of the most commonly used oral treatments based on WHO guidelines for treatment of COVID-19, HIV, HBV and HCV. Current lowest global prices were obtained from public reports and the Pan American Health Organization Antiretroviral Report. Our target prices were compared with national pricing data from a range of low, middle, and high-income countries. Results The main results table shows current prices of antiretrovirals for SARS-CoV-2 (5–14-day course), HIV or HBV (per 365-day course), and HCV treatments (per 12-week course). COVID-19 can be treated with molnupiravir for $12.40, baricitinib for $0.60, or dexamethasone for $0.19 per course. HIV can be treated with DRV/r for $286 per year, ATV/r for $140 per year, TDF/3TC/DTG for $74 per year, TAF/FTC/DTG for $89 and TDF/FTC for $59. HBV can be treated with TDF for $23 per year and TAF for $38 per year. HCV could be cured with sofosbuvir/daclatasvir for $35 per patient and sofosbuvir/velpatasvir for $79 per patient. Maximum list prices (typically from US) were up to 600 times higher than costs of production (e.g. for TDF). Conclusions Key viral infections can be treated or cured with generic drugs at prices far below those of branded equivalents. Use of branded drugs at high prices can limit the potential for countries to achieve the UNAIDS 95-95-95 targets.

Background Tobacco smoking is estimated to account for more than 1 million annual deaths in China, and the epidemic continues to increase in men. Large nationwide prospective studies linked to different health records can help to periodically assess disease burden attributed to smoking. We aimed to examine associations of smoking with incidence of and mortality from an extensive range of diseases in China. Methods We analysed data from the prospective China Kadoorie Biobank, which recruited 512 726 adults aged 30–79 years, of whom 210 201 were men and 302 525 were women. Participants who had no major disabilities were identified through local residential records in 100–150 administrative units, which were randomly selected by use of multistage cluster sampling, from each of the ten diverse study areas of China. They were invited and recruited between June 25, 2004, and July 15, 2008. Upon study entry, trained health workers administered a questionnaire assessing detailed smoking behaviours and other key characteristics (eg, sociodemographics, lifestyle, and medical history). Participants were followed up via electronic record linkages to death and disease registries and health insurance databases, from baseline to Jan 1, 2018. During a median 11-year follow-up (IQR 10–12), 285 542 (55·7%) participants were ever hospitalised, 48 869 (9·5%) died, and 5252 (1·0%) were lost to follow-up during the age-at-risk of 35–84 years. Cox regression yielded hazard ratios (HRs) associating smoking with disease incidence and mortality, adjusting for multiple testing. Findings At baseline, 74·3% of men and 3·2% of women (overall 32·4%) ever smoked regularly. During follow-up, 1 137 603 International Classification of Diseases, 10th revision (ICD-10)-coded incident events occurred, involving 476 distinct conditions and 85 causes of death, each with at least 100 cases. Compared with never-regular smokers, ever-regular smokers had significantly higher risks for nine of 18 ICD-10 chapters examined at age-at-risk of 35–84 years. For individual conditions, smokers had significantly higher risks of 56 diseases (50 for men and 24 for women) and 22 causes of death (17 for men and nine for women). Among men, ever-regular smokers had an HR of 1·09 (95% CI 1·08–1·11) for any disease incidence when compared with never-regular smokers, and significantly more episodes and longer duration of hospitalisation, particularly those due to cancer and respiratory diseases. For overall mortality, the HRs were greater in men from urban areas than in men from rural areas (1·50 [1·42–1·58] vs 1·25 [1·20–1·30]). Among men from urban areas who began smoking at younger than 18 years, the HRs were 2·06 (1·89–2·24) for overall mortality and 1·32 (1·27–1·37) for any disease incidence. In this population, 19·6% of male (24·3% of men residing in urban settings and 16·2% of men residing in rural settings) and 2·8% of female deaths were attributed to ever-regular smoking. Interpretation Among Chinese adults, smoking was associated with higher risks of morbidity and mortality from a wide range of diseases. Among men, the future smoking-attributed disease burden will increase further, highlighting a pressing need for reducing consumption through widespread cessation and uptake prevention. Funding British Heart Foundation, Cancer Research UK, Chinese Ministry of Science and Technology, Kadoorie Charitable Foundation, UK Medical Research Council, National Natural Science Foundation of China, Wellcome Trust.

This brief proposes a quasi time-fuel optimal control strategy to solve the dynamic tracking problem of unmanned systems when fuel and control input are limited. This kind of motion planning and control strategy could bring the biggest advantage into full play in the field of switching tracking for multiple targets, such as the multi-target strike of weapons and the rapid multi-target grabbing of robots on industrial assembly lines. Compared with the time-fuel optimal control that has been studied before, the proposed controller retains the advantages of optimal control when switching between multiple dynamic targets, and improves the high-frequency oscillation problem of the system under the discontinuous control strategy. Moreover, the asymmetry of friction load, which can affect the dynamic performance of the system, is also considered in this brief. Therefore, the novel control law, which is addressed by this brief, can make the corresponding system achieve the desired transient performance and satisfactory steady-state performance when switching between multiple dynamic targets. The experiment results based on the visual tracking turntable are presented to verify the superiority of the proposed method.

We report a facile method for preparation of amphiphilic Janus silica nanospheres using layered double hydroxide as a protective support. The unprotected part of silica nanospheres is modified with...

The accuracy of trajectory tracking and stable operation with heavy load are the main challenges of parallel mechanism for wheel-legged robots, especially in complex road conditions. To guarantee the tracking performance in an uncertain environment, the disturbances, including the internal-robot friction and external-robot and environment interaction forces, should be considered in the robot’s dynamical system. In this article, a neural fuzzy-based model predictive tracking scheme (NFMPC) for reliable tracking control is proposed to the developed four wheel-legged robot, and the fuzzy neural network approximation is applied to estimate the unknown physical interaction and external dynamics of the robot system. Meanwhile, the advanced parallel mechanism of the four wheel-legged robot (BIT-NAZA) is introduced. Finally, co-simulation and experiment results using the BIT-NAZA robot derived from the proposed hybrid control strategy indicate that the methodology can achieve satisfactory tracking performance in terms of accuracy and stability. This research can provide theoretical and engineering guidance for lateral stability of intelligent robots under unknown disturbances and uncertain nonlinearities, and facilitate the control performance of the wheel-legged robot in a practical system.

This paper proposes a neural approximation based model predictive control approach for tracking control of a nonholonomic wheel-legged robot in complex environments, which features mechanical model uncertainty and unknown disturbances. In order to guarantee the tracking performance of wheel-legged robots in an uncertain environment, effective approaches for reliable tracking control should be investigated with the consideration of the disturbances, including internal-robot friction and external physical interactions in the robot’s dynamical system. In this paper, a radial basis function neural network (RBFNN) approximation based model predictive controller (NMPC) is designed and employed to improve the tracking performance for nonholonomic wheel-legged robots. Some demonstrations using a BIT-NAZA robot are performed to illustrate the performance of the proposed hybrid control strategy. The results indicate that the proposed methodology can achieve promising tracking performance in terms of accuracy and stability.

In order to realise autonomous navigation of unmanned platforms in urban or off-road environments, it is crucial to study accurate, versatile and real-time road detection methods. This study proposes an adaptive road detection method that combines lane lines and obstacle boundaries, which can be applied to a variety of driving environments. Combining multi-channel threshold processing, it is robust to lane feature detection under various complex situations. Obstacle information extracted from the grid image constructed by 3D LIDAR point cloud is used for lane feature selection to avoid interference from pedestrians and vehicles. The proposed method makes use of adaptive sliding window for feature selection, and piecewise least squares method for road line fitting. Experimental results on dataset and in real-world environments show that the proposed method can overcome illumination changes, shadow occlusion, pedestrian, vehicle interference and so on in a variety of scenes. The proposed method has good enough efficiency, robustness and real-time performance.