Speech comprehension is the key to Human-Robot Interaction (HRI), however, noise energetic masking can greatly reduce the performance of speech comprehension. Simply using Speech Enhancement (SE) methods can easily lead to suboptimal solutions. To alleviate these issues, this paper proposes a Noise-Robust Speech Comprehension system based on Knowledge Graph (NRSC-KG), which designs robust methods in Automatic Speech Recognition (ASR) and ASR error correction. In ASR, we propose a Self-adaptive Speech Feature Fusion Network (SSFF-Net) to dynamically fuse enhanced speech features with the original speech features. In ASR error correction, we propose a divergent apperceive error correction method based on domain-specific Knowledge Graph (KG). The experiment on AISHELL-1 speech corpus shows that compared to the traditional joint training of SE and ASR, our method achieves a relative Character Error Rate (CER) reduction of 12.7%. We validated the KG based error correction method on our customized Herbal Cuisine ordering Dialogue speech dataset (HCDialogue), and the results showed that our method achieved a relative CER reduction of 5.0% and 2.1%, respectively, compared to the language model based and end-to-end neural network based methods. It’s indicated that the KG based method has great potential in improving speech comprehension performance.

Chili recognition is one of the critical technologies for robots to pick chilies. The robots need locate the fruit. Furthermore, chilies are always planted intensively and their fruits are always clustered. It is a challenge to recognize and locate the chilies that are blocked by branches and leaves, or other chilies. However, little is known about the recognition algorithms considering this situation. Failure to solve this problem will mean that the robot cannot accurately locate and collect chilies, which may even damage the picking robot’s mechanical arm and end effector. Additionally, most of the existing ground target recognition algorithms are relatively complex, and there are many problems, such as numerous parameters and calculations. Many of the existing models have high requirements for hardware and poor portability. It is very difficult to perform these algorithms if the picking robots have limited computing and battery power. In view of these practical issues, we propose a target recognition-location scheme GNPD-YOLOv5s based on improved YOLOv5s in order to automatically identify the occluded and non-occluded chilies. Firstly, the lightweight optimization for Ghost module is introduced into our scheme. Secondly, pruning and distilling the model is designed to further reduce the number of parameters. Finally, the experimental data show that compared with the YOLOv5s model, the floating point operation number of the GNPD-YOLOv5s scheme is reduced by 40.9%, the model size is reduced by 46.6%, and the reasoning speed is accelerated from 29 ms/frame to 14 ms/frame. At the same time, the Mean Accuracy Precision (MAP) is reduced by 1.3%. Our model implements a lightweight network model and target recognition in the dense environment at a small cost. In our locating experiments, the maximum depth locating chili error is 1.84 mm, which meets the needs of a chili picking robot for chili recognition.

Autonomous mobile service robots are used to complete many tasks, such as cleaning, transporting goods and monitoring. Such tasks usually require uninterrupted and continuous service. However, the battery of the robot is limited and must be charged frequently. For a large number of robots, it is essential to select a suitable charging pile. For this issue, we propose a path planning model for robots to intelligently access a limited number of charging piles distributed on the map. The traditional path planning model mainly considers the shortest path criterion to generate the path. Different from this, the path planning model in this paper not only considers the shortest path, but also the service position of the robot after charging, the remaining power of the robot, the state of the charging pile and the position of the robot in the map. Our path planning assigns the most suitable charging pile to the robot that needs charging. To solve the problem of high memory consumption and slow search speed when traditional $\text{A}^{\ast} $ algorithm is used for path planning, we propose local memorial path planning (LMPP) algorithm to quickly generate effective paths. The simulation results show that the proposed robot charging path planner can improve the robot service satisfaction and plan the effective path to the available charging piles.

Path planning is a challenging, computationally complex optimization task in high-dimensional scenarios. The metaheuristic algorithm provides an excellent solution to this problem. The dung beetle optimizer (DBO) is a recently developed metaheuristic algorithm inspired by the biological behavior of dung beetles. However, it still has the drawbacks of poor global search ability and being prone to falling into local optima. This paper presents a multi-strategy enhanced dung beetle optimizer (MDBO) for the three-dimensional path planning of an unmanned aerial vehicle (UAV). First, we used the Beta distribution to dynamically generate reflection solutions to explore more search space and allow particles to jump out of the local optima. Second, the Levy distribution was introduced to handle out-of-bounds particles. Third, two different cross operators were used to improve the updating stage of thief beetles. This strategy accelerates convergence and balances exploration and development capabilities. Furthermore, the MDBO was proven to be effective by comparing seven state-of-the-art algorithms on 12 benchmark functions, the Wilcoxon rank sum test, and the CEC 2021 test suite. In addition, the time complexity of the algorithm was also analyzed. Finally, the performance of the MDBO in path planning was verified in the three-dimensional path planning of UAVs in oil and gas plants. In the most challenging task scenario, the MDBO successfully searched for feasible paths with the mean and standard deviation of the objective function as low as 97.3 and 32.8, which were reduced by 39.7 and 14, respectively, compared to the original DBO. The results demonstrate that the proposed MDBO had improved optimization accuracy and stability and could better find a safe and optimal path in most scenarios than the other metaheuristics.

It is necessary for robot-trailer to tow robot to the charging station for recharging when the robot fails, or the battery cannot support the robot to the charging station. However, the delay of towing robot is affecting the working efficiency of mobile robot. Based on the task priority of the mobile robot and impact degree on the room after the power failure, the paper proposes a distributed scheduling of robot being towed to recharge for reducing the delay expectation. This work designs a Distributed Three Nodes Service (DTNS) scheduling based on the communication between charging stations. The two-step path-planner based on decision factor and travel path is used in the scheme. Simulations show that the distributed scheduling of this work can well ensure the success communication in the case of low power, and DTNS can well reduce the delay expectation of towing robot to recharge. Compared with First Come First Service (FCFS) scheduling, DTNS reduces the towing delay by 48.71%, 48.83% and 40.45% when there are some robots sending the towing request, and by 58.77%, 39.97% and 41.90% when no robot sends request in the case of 1, 2 and 3 robot trailers in the service space respectively.

License plate (LP) detection is a crucial task for Automatic License Plate Recognition (ALPR) systems. Most existing LP detection networks can detect License plates, but their accuracy suffers when license plates (LPs) are tilted or deformed due to perspective distortion. This is because these detectors can only detect the region where the LP is located, and even the most advanced object detectors struggle in unconstrained scenarios. To address this problem, we propose a lightweight Deformation Planar Object Detection Network (DPOD-NET), which can correct the deformed LPs of various vehicles (e.g., car, truck, electric motorcycle, bus) by detecting the LP corner points. Accordingly, the distortion associated with perspective is mitigated when we adjust the LP to a frontal parallel view through the LP corners. To optimize small errors between the predicted and true values of the LP corner points, we propose an LPWing loss function. Compared with the commonly used L1 function, the LPWing loss is derivable at the zero position, and the gradient will be bigger when errors are smaller. This enables the model to converge faster at the position where the error is close to zero, resulting in better convergence when the error between the true values and predicted values is small. In addition, the paper presents a stochastic multi-scale image detail boosting strategy, which effectively augments the dataset. Finally, to objectively evaluate the effectiveness of LP corner detection approaches, we present a dataset (LPDE-4K) including various LP types (e.g., color, country, illumination, distortion). We test the performance on various datasets, and our approach outperforms other existing state-of-the-art approaches in terms of higher accuracy and lower computational cost.

Aiming at the problems of high risk and low efficiency in the process of transportation of chemical materials, which are corrosive, toxic and harmful, flammable, and explosive in the modern intelligent warehousing logistics and supply chain system, a universal dual-side butt automatic flip machine was designed, and the main structure and working principle of the flip machine were introduced. Through the theoretical calculation and virtual simulation, the structural design of the flip machine was completed, and it was concluded that the deformation and stress of the key parts of the flip machine, such as the clamping mechanism, main rotating shaft, and support frame, meet the design requirements under 1.35t rated external load. On this basis, the prototype of the flip machine was made, and 2880 times of flipping tests were carried out in three days. The results show that the flip machine has a remarkable level of safety and a high degree of automation, and the flipping efficiency of materials under 1.35t can reach 40 pallets/h, which is greatly improved compared with the traditional mode, and the feasibility and reliability were preliminarily verified.

Many recipes contain ingredients with various anticancer effects, which can help users to prevent cancer, as well as provide treatment for cancer patients, effectively slowing the disease. Existing recipe knowledge graph recommendation systems obtain entity feature representations by mining latent connections between recipes and between users and recipes to enhance the performance of the recommendation system. However, it ignores the influence of time on user taste preferences, fails to capture the dependency between them from the user’s dietary records, and is unable to more accurately predict the user’s future recipes. We use the KGAT to obtain the embedding representation of entities, considering the influence of time on users, and recipe recommendation can be viewed as a long-term sequence prediction, introducing LSTM networks to dynamically adjust users’ personal taste preferences. Based on the user’s dietary records, we infer the user’s preference for the future diet. Combined with the cancer knowledge graph, we provide the user with diet recommendations that are beneficial to disease prevention and rehabilitation. To verify the effectiveness and rationality of PPKG, we compared it with three other recommendation algorithms on the self-created datasets, and the extensive experimental results demonstrate that our algorithm performance performs other algorithms, which confirmed the effectiveness of PPKG in dealing with sequence recommendation.

Tea picking has always been mainly manual, but with the development of technology, it has been possible to use computer image recognition technology to assist robots in identifying tea leaves and picking them. However, the special characteristics of tea leaves make it difficult for the machine to recognize them with high accuracy and efficiency. In this paper, we compare the currently available algorithms and perform tea image recognition based on the YOLOv5 algorithm model. Based on the original YOLOv5 model, the CBAM attention mechanism is added to improve the recognition ability of tea. The mAP of our model can reach 89.1%, which is 13.6 percentage points higher than the original YOLOv5 model.

针对相对密闭的场所等特殊环境对室内灯光管理的需求,提出了一种新的智能灯控方法,并设计了一种以ZigBee为核心的人脸识别灯控制系统.该系统首先通过ZigBee节点采集并发送室内温湿度、异常气体等信息并上传到手机端和服务器端,通过提出的基于ArcFace的自然场景人脸识别算法识别终端采集的特定人脸,从而实现对终端灯光及数据的控制.系统测试实验表明,该系统设计简单、成本低,采用人脸识别加密,提高了灯控的安全性,具有较为广泛的应用价值.

In mobile data acquisition, mobile robots usually face challenging tasks when collecting information in an undetermined environment with energy limitation and time-sensitive requirements. We formulate the task of data acquisition as a multiobjective optimization problem under energy and time constraints. In our investigation, three objectives for data acquisition are considered, including collecting the largest amount of information, moving along a path with the smallest probability of encountering obstacles, and traveling with shortest possible overall distance. To resolve the formulated problem which yields the best path for a mobile robot, we propose a mixed cognition particle swarm optimization (MCPSO) algorithm, which adopts the min-max normalization to calculate the fitness, and we transform the multiobjective optimization problem into a single-objective optimization problem by summation after normalization. The efficiency of the MCPSO algorithm is evaluated for mobile data acquisition in several well-known benchmarks by simulation. The simulation results demonstrate that the proposed MCPSO algorithm can achieve higher accuracy and faster convergence compared with other particle swarm optimization algorithms.

In the smart mariculture, the timely and accurate predictions of water quality can help farmers take countermeasures before the ecological environment deteriorates seriously. However, the openness of the mariculture environment makes the variation of water quality nonlinear, dynamic and complex. Traditional methods face challenges in prediction accuracy and generalization performance. To address these problems, an accurate water quality prediction scheme is proposed for pH, water temperature and dissolved oxygen. First, we construct a new huge raw data set collected in time series consisting of 23,204 groups of data. Then, the water quality parameters are preprocessed for data cleaning successively through threshold processing, mean proximity method, wavelet filter, and improved smoothing method. Next, the correlation between the water quality to be predicted and other dynamics parameters is revealed by the Pearson correlation coefficient method. Meanwhile, the data for training is weighted by the discovered correlation coefficients. Finally, by adding a backward SRU node to the training sequence, which can be integrated into the future context information, the deep Bi-S-SRU (Bi-directional Stacked Simple Recurrent Unit) learning network is proposed. After training, the prediction model can be obtained. The experimental results demonstrate that our proposed prediction method achieve higher prediction accuracy than the method based on RNN (Recurrent Neural Network) or LSTM (Long Short-Term Memory) with similar or less time computing complexity. In our experiments, the proposed method takes 12.5ms to predict data on average, and the prediction accuracy can reach 94.42% in the next 3~8 days.

An accurate prediction of cage-cultured water quality is a hot topic in smart mariculture. Since the mariculturing environment is always open to its surroundings, the changes in water quality parameters are normally nonlinear, dynamic, changeable, and complex. However, traditional forecasting methods have lots of problems, such as low accuracy, poor generalization, and high time complexity. In order to solve these shortcomings, a novel water quality prediction method based on the deep LSTM (long short-term memory) learning network is proposed to predict pH and water temperature. Firstly, linear interpolation, smoothing, and moving average filtering techniques are used to repair, correct, and de-noise water quality data, respectively. Secondly, Pearson’s correlation coefficient is used to obtain the correlation priors between pH, water temperature, and other water quality parameters. Finally, a water quality prediction model based on LSTM is constructed using the preprocessed data and its correlation information. Experimental results show that, in the short-term prediction, the prediction accuracy of pH and water temperature can reach 98.56% and 98.97%, and the time cost of the predictions is 0.273 s and 0.257 s, respectively. In the long-term prediction, the prediction accuracy of pH and water temperature can reach 95.76% and 96.88%, respectively.

Data acquisition in large areas has issues of cost and data loss. When sensors are sparse in the physical field, it is critical to study the deployment methods to improve the accuracy of reconstructed data set and the precision of the recovery of lost data. It is desirable to place sensors at optimal locations to achieve higher precision of recovery. In this paper, we present a sparse sensor placement scheme for data interpolation reconstruction based on iterative four subregions using fractal theory. The results of our experiments demonstrate that the precision of our algorithm is higher than that with random placement in dispersion degree, coverage rate, and reconstruction accuracy.

The sampling rate of wideband spectrum sensing for sparse signals can be reduced by sub-Nyquist sampling with a Modulated Wideband Converter (MWC). In collaborative spectrum sensing, the fusion center recovers the spectral support from observation and measurement matrices reported by a network of CRs, to improve the precision of spectrum sensing. However, the MWC has a very high hardware complexity due to its parallel structure; it sets a fixed threshold for a decision without considering the impact of noise intensity, and needs a priori information of signal sparsity order for signal support recovery. To address these shortcomings, we propose a progressive support selection based self-adaptive distributed MWC sensing scheme (PSS-SaDMWC). In the proposed scheme, the parallel hardware sensing channels are scattered on secondary users (SUs), and the PSS-SaDMWC scheme takes sparsity order estimation, noise intensity, and transmission loss into account in the fusion center. More importantly, the proposed scheme uses a support selection strategy based on a progressive operation to reduce missed detection probability under low SNR levels. Numerical simulations demonstrate that, compared with the traditional support selection schemes, our proposed scheme can achieve a higher support recovery success rate, lower sampling rate, and stronger time-varying support recovery ability without increasing hardware complexity.

During the construction of wireless sensor networks (WSNs) for smart cities, a preliminary survey of the relative criticalness within the monitored area can be performed. It is a challenge for deterministic sensor deployment to balance the tradeoff of sensing reliability and cost. In this paper, based on the sensing accuracy of the sensor, we establish a reliability model of the sensing area which is divided into sensing grids, and different weights are allocated to those grids. We employ a practical evaluation criterion using seesaw mapping for determining the weights of sensing grids. We further formulate and solve an optimization problem for maximizing the trust degree of the WSNs. With our proposed method, the efficient deployment of sensors can be realized. Simulation results show that our proposed deployment strategy can achieve higher trust degree with reduced sensor deployment cost and lower number of sensors at a certain miss probability threshold.

The data space collected by a wireless sensor network (WSN) is the basis of data mining and data visualization. In the process of monitoring physical quantities with large time and space correlations, incomplete acquisition strategy with data interpolation can be adopted to reduce the deployment cost. To improve the performance of data interpolation in such a scenario, we proposed a robust data interpolation based on a back propagation artificial neural network operator. In this paper, a neural network learning operator is proposed based on the strong fault tolerance of artificial neural networks. The learning operator is trained by using the historical data of the data acquisition nodes of WSN and is transferred to estimate the value of physical quantities at the locations where sensors are not deployed. The experimental results show that our proposed method yields smaller interpolation error than the traditional inverse-distance-weighted interpolation (IDWI) method.

Spectrum sensing is one of the key technologies in wireless wideband communication. There are still challenges in respect of how to realize fast and robust wideband spectrum sensing technology. In this paper, a novel nonreconstructed sequential compressed wideband spectrum sensing algorithm (NSCWSS) is proposed. Firstly, the algorithm uses a sequential spectrum sensing method based on history memory and reputation to ensure the robustness of the algorithm. Secondly, the algorithm uses the strategy of compressed sensing without reconstruction, which thus ensures the sensing agility of the algorithm. The algorithm is simulated and analyzed by using the centralized cooperative sensing. The theoretical analysis and simulation results reveal that, under the condition of ensuring the certain detection probability, the proposed algorithm effectively reduces complex computation of signal reconstruction, significantly reducing the wideband spectrum sampling rate. At the same time, in the cognitive wideband communication scenarios, the algorithm also achieves a better defense against the SSDF attack in spectrum sensing.

As a vital issue of information based fisheries,morphological feature detection of fish is widely applied to many domains. The existing fish eye feature detection mainly depends on manual work,and the operation process is complex with low efficiency and subjective results. Therefore,a detection method of fish eye based on image processing and least square ellipse fitting is proposed. Firstly,fish images are collected from facilities in farm. Then image preprocessing is performed on fish eye pupil and iris. Finally,the edge of the fish eye pupil and iris is identified by ellipse fitting method. As seen from the results,pupil detection costs an average time of 325. 96 ms,iris detection costs an average time of 364. 57 ms,pupil detection error is7. 247( 1±11.82%) mm,and iris detection error is 12.179( 1 ± 14.05%) mm. Studies have shown that the proposed method can not only solve the problem that manual work is complex and subjective,but also avoid the time-consuming measuring process and provide a new way for large quantities of fish eye non-contact measurement.

Privacy-preserving in wireless sensor networks is one of the key problems to be solved in practical applications. It is of great significance to solve the problem of data privacy protection for large-scale applications of wireless sensor networks. The characteristics of wireless sensor networks make data privacy protection technology face serious challenges. At present, the technology of data privacy protection in wireless sensor networks has become a hot research topic, mainly for data aggregation, data query, and access control of data privacy protection. In this paper, multiorder fusion data privacy-preserving scheme (MOFDAP) is proposed. Random interference code, random decomposition of function library, and cryptographic vector are introduced for our proposed scheme. In multiple stages and multiple aspects, the difficulty of cracking and crack costs are increased. The simulation results demonstrate that, compared with the typical Slice-Mix-AggRegaTe (SMART) algorithm, the algorithm proposed in this paper has a better data privacy-preserving ability when the traffic load is not very heavy.