Search results

Intent detection (ID) and slot filling (SF) are two important tasks in natural language understanding. ID is to identify the main intent of a paragraph of text. The goal of SF is to extract the information that is important to the intent from the input sentence. However, most of the existing methods use sentence-level intention recognition, which has the risk of error propagation, and the relationship between intention recognition and SF is not explicitly modeled. Aiming at this problem, this paper proposes a collaborative model of ID and SF for intelligent spoken language understanding called ID-SF-Fusion.

ID-SF-Fusion uses Bidirectional Encoder Representation from Transformers (BERT) and Bidirectional Long Short-Term Memory (BiLSTM) to extract effective word embedding and context vectors containing the whole sentence information respectively. Fusion layer is used to provide intent–slot fusion information for SF task. In this way, the relationship between ID and SF task is fully explicitly modeled. This layer takes the result of ID and slot context vectors as input to obtain the fusion information which contains both ID result and slot information. Meanwhile, to further reduce error propagation, we use word-level ID for the ID-SF-Fusion model. Finally, two tasks of ID and SF are realized by joint optimization training.

We conducted experiments on two public datasets, Airline Travel Information Systems (ATIS) and Snips. The results show that the Intent ACC score and Slot F1 score of ID-SF-Fusion on ATIS and Snips are 98.0 per cent and 95.8 per cent, respectively, and the two indicators on Snips dataset are 98.6 per cent and 96.7 per cent, respectively. These models are superior to slot-gated, SF-ID NetWork, stack-Prop and other models. In addition, ablation experiments were performed to further analyze and discuss the proposed model.

This paper uses word-level intent recognition and introduces intent information into the SF process, which is a significant improvement on both data sets.

Intent detection and slot filling are two important tasks in question comprehension of a question answering system. This study aims to build a joint task model with some generalization ability and benchmark its performance over other neural network models mentioned in this paper.

This study used a deep-learning-based approach for the joint modeling of question intent detection and slot filling. Meanwhile, the internal cell structure of the long short-term memory (LSTM) network was improved. Furthermore, the dataset Computer Science Literature Question (CSLQ) was constructed based on the Science and Technology Knowledge Graph. The datasets Airline Travel Information Systems, Snips (a natural language processing dataset of the consumer intent engine collected by Snips) and CSLQ were used for the empirical analysis. The accuracy of intent detection and F1 score of slot filling, as well as the semantic accuracy of sentences, were compared for several models.

The results showed that the proposed model outperformed all other benchmark methods, especially for the CSLQ dataset. This proves that the design of this study improved the comprehensive performance and generalization ability of the model to some extent.

This study contributes to the understanding of question sentences in a specific domain. LSTM was improved, and a computer literature domain dataset was constructed herein. This will lay the data and model foundation for the future construction of a computer literature question answering system.

Flights are often delayed owing to emergencies. This paper proposes a cooperative slot secondary assignment (CSSA) model based on a collaborative decision-making (CDM) mechanism, and the operation mode of flight waves designs an improved intelligent algorithm to solve the optimal flight plan and minimize the total delay of passenger time.

Taking passenger delays, transfer delays and flight cancellation delays into account comprehensively, the total delay time is minimized as the objective function. The model is verified by a linear solver and compared with the first come first service (FCFS) method to prove the effectiveness of the method. An improved adaptive partheno-genetic algorithm (IAPGA) using hierarchical serial number coding was designed, combining elite and roulette strategies to find pareto solutions.

Comparing and analyzing the experimental results of various scale examples, the optimization model in this paper is greatly optimized compared to the FCFS method in terms of total delay time, and the IAPGA algorithm is better than the algorithm before in terms of solution performance and solution set quality.

Based on the actual situation, this paper considers the operation mode of flight waves. In addition, the flight plan solved by the model can be guaranteed in terms of feasibility and effectiveness, which can provide airlines with reasonable decision-making opinions when reassigning slot resources.

This paper aims to present the simulation, fabrication and testing of a novel ultra-wide band (UWB) band-pass filters (BPFs) with better transmission and rejection characteristics on a low-loss Taconic substrate and analyze using the coupled theory of resonators for UWB range covering L, S, C and X bands for radars, global positioning system (GPS) and satellite communication applications.

The filter is designed with a bent coupled transmission line on the top copper layer. Defected ground structures (DGSs) like complementary split ring resonators (CSRRs), V-shaped resonators, rectangular slots and quad circle slots (positioned inwards and outwards) are etched in the ground layer of the filter. The circular orientation of V-shaped resonators adds compactness when linearly placed. By arranging the quad circle slots outwards and inwards at the corner and core of the ground plane, respectively, two filters (Filters I and II) are designed, fabricated and measured. These two filters feature a quasi-elliptic response with transmission zeros (TZs) on either side of the bandpass response, making it highly selective and reflection poles (RPs), resulting in a low-loss filter response. The transmission line model and coupled line theory are implemented to analyze the proposed filters.

Two filters by placing the quad circle slots outwards (Filter I) and inwards (Filter II) were designed, fabricated and tested. The fabricated model (Filter I) provides transmission with a maximum insertion loss of 2.65 dB from 1.5 GHz to 9.2 GHz. Four TZs and five RPs are observed in the frequency response. The lower and upper stopband band width (BW) of the measured Filter I are 1.2 GHz and 5.5 GHz of upper stopband BW with rejection level greater than 10 dB, respectively. Filter II (inward quad circle slots) operates from 1.4 GHz to 9.05 GHz with 1.65 dB maximum insertion loss inside the passband with four TZs and four RPs, which, in turn, enhances the filter characteristics in terms of selectivity, flatness and stopband. Moreover, 1 GHz BW of lower and upper stopbands are observed. Thus, the fabricated filters (Filters I and II) are therefore evaluated, and the outcomes show good agreement with the electromagnetic simulation response.

The limitation of this work is the back radiation caused by DGS, which can be eradicated by placing the filter in the cavity and retaining its performance.

The proposed UWB BPFs with novel resonators find their role in the UWB range covering L, S, C and X bands for radars, GPS and satellite communication applications.

To the best of the authors’ knowledge, for the first time, the authors develop a compact UWB BPFs (Filters I and II) with BW greater than 7.5 GHz by combining reformed coupled lines and DGS resonators (CSRRs, V-shaped resonators [modified hairpin resonators], rectangular slots and quad circle slots [inwards and outwards]) for radars, GPS and satellite communication applications.

This study addresses the production optimization of a cellular manufacturing system (CMS) in magnetic production enterprises. Magnetic products and raw materials are more critical to transport than general products because the attraction or repulsion between magnetic poles can easily cause traffic jams. This study needs to address a method to promote the scheduling efficiency of the problem.

To address this problem, this study formulated a mixed-integer linear programming (MILP) model to describe the problem and proposed an auction and negotiation-based approach with a local search to solve it. Auction- and negotiation-based approaches can obtain feasible and high-quality solutions. A local search operator was proposed to optimize the feasible solutions using an improved conjunctive graph model.

Verification tests were performed on a series of numerical examples. The results demonstrated that the proposed auction and negotiation-based approach with a local search operator is better than existing solution methods for the problem identified. Statistical analysis of the experiment results using the Statistical Package for the Social Sciences (SPSS) software demonstrated that the proposed approach is efficient, stable and suitable for solving large-scale numerical instances.

An improved auction and negotiation-based approach was proposed; The conjunctive graph model was also improved to describe the problem of CMS with traffic jam constraint and build the local search operator; The authors’ proposed approach can get better solution than the existing algorithms by testing benchmark instances and real-world instances from enterprises.

This paper aims to present the design development and measurement of two aerodynamic slotted X-bands back-to-back planer substrate-integrated rectangular waveguide (SIRWG/SIW) to Microstrip (MS) line transition for satellite and RADAR applications. It facilitates the realization of nonplanar (waveguide-based) circuits into planar form for easy integration with other planar (microstrip) devices, circuits and systems. This paper describes the design of a SIW to microstrip transition. The transition is broadband covering the frequency range of 8–12 GHz. The design and interconnection of microwave components like filters, power dividers, resonators, satellite dishes, sensors, transmitters and transponders are further aided by these transitions. A common planar interconnect is designed with better reflection coefficient/return loss (RL) (S₁₁/S₂₂ ≤ 10 dB), transmission coefficient/insertion loss (IL) (S₁₂/S₂₁: 0–3.0 dB) and ultra-wideband bandwidth on low profile FR-4 substrate for X-band and Ku-band functioning to interconnect modern era MIC/MMIC circuits, components and devices.

Two series of metal via (6 via/row) have been used so that all surface current and electric field vectors are confined within the metallic via-wall in SIW length. Introduced aerodynamic slots in tapered portions achieve excellent impedance matching and tapered junctions with SIW are mitered for fine tuning to achieve minimum reflections and improved transmissions at X-band center frequency.

Using this method, the measured IL and RLs are found in concord with simulated results in full X-band (8.22–12.4 GHz). RLC T-equivalent and p-equivalent electrical circuits of the proposed design are presented at the end.

The measurement of the prototype has been carried out by an available low-cost X-band microwave bench and with a Keysight E4416A power meter in the microwave laboratory.

The transition is fabricated on FR-4 substrate with compact size 14 mm × 21.35 mm × 1.6 mm and hence economical with IL lie within limits 0.6–1 dB and RL is lower than −10 dB in bandwidth 7.05–17.10 GHz. Because of such outstanding fractional bandwidth (FBW: 100.5%), the transition could also be useful for Ku-band with IL close to 1.6 dB.

Data traffic through wireless communication is significantly increasing, resulting in the frequency of streaming applications as various formats and the evolution of the Internet of Things (IoT), such as virtual reality, edge device based transportation and surveillance systems. Growth in kind of applications resulted in increasing the scope of wireless communication and allocating a spectrum, as well as methods to decrease the intervention between nearby-located wireless links functioning on the same spectrum bands and hence to proliferation for the spectral efficiency. Recent advancement in drone technology has evolved quickly leading on board sensors with increased energy, storage, communication and processing capabilities. In future, the drone sensor networks will be more common and energy utilization will play a crucial role to maintain a fully functional network for the longest period of time. Envisioning the aerial drone network, this study proposes a robust high level design of algorithms for the drones (group coordination). The proposed design is validated with two algorithms using multiple drones consisting of various on-board sensors. In addition, this paper also discusses the challenges involved in designing solutions. The result obtained through proposed method outperforms the traditional techniques with the transfer rate of more than 3 MB for data transfer in the drone with coordination

Fair Scheduling Algorithm (FSA) using a queue is a distributed slot assignment algorithm. The FSA executes in rounds. The duration of each round is dynamic based upon the delay in the network. FSA prevents the collision by ensuring that none of the neighboring node gets the same slot. Nodes (Arivudainambi et al., 2019) which are separated by two or more hopes can get assigned in the same slot, thereby preventing the collision. To achieve fairness at the scheduling level, the FSA maintains four different states for each node as IDLE, REQUEST, GRANT and RELEASE.

A multi-unmanned aerial vehicle (UAV) system can operate in both centralized and decentralized manner. In a centralized system, the ground control system will take care of drone data collection, decisions on navigation, task updation, etc. In a decentralized system, the UAVs are unambiguously collaborating on various levels as mentioned in the centralized system to achieve the goal which is represented in Figure 2.

However, the multi-UAVs are context aware in situations such as environmental observation, UAV–UAV communication and decision-making. Independent of whether operation is centralized or decentralized, this study relates the goals of the multi-UAVs are sensing, communication and coordination among other UAVs, etc. Figure 3 shows overall system architecture.

The individual events attempts in the UAV’s execution are required to complete the mission in superlative manner which affects in every multi UAV system. This multi UAV systems need to take a steady resolute on what way UAV has to travel and what they need to complete to face the critical situations in changing of environments with the uncertain information. This coordination algorithm has certain dimensions including events that they needs to resolute on, the information that they used to make a resolution, the resolute making algorithm, the degree of decentralization. In multi UAV systems, the coordinated events ranges from lower motion level.

This study has proposed a novel self-organizing coordination algorithm for multi-UAV systems. Further, the experimental results also confirm that is robust to form network at ease. The testbed for this simulation to sensing, communication, evaluation and networking. The algorithm coordination has to testbed with multi UAVs systems. The two scheduling techniques has been used to transfer the packets using done network. The self-organizing algorithm (SOA) with fair scheduling queue outperforms the weighted queue scheduling in the transfer rate with less loss and time lag. The results obtained through from Figure 10 clearly indicates that the fair queue scheduling with SOA have several advantages over weighted fair queue in different parameters.

Motivated by the real-world practice that the boom of the online selling induces a higher product return as well, selecting which online channel mode indicates who takes ownership over the product and thus bears the loss of the product return. This paper aims to seek the optimal online channel modes for the two members in a platform supply chain in the presence of product returns.

This study aims to develop a platform supply chain that consists of one platform company and one supplier. Along with an offline distribution channel, the supplier can choose two alternative online selling modes (i.e. the reselling and agency modes) to sell its product through the online marketplace. This paper applies Stackelberg game to derive the equilibrium with different business scenarios and selects the optimal online channel modes for two parties, respectively. Moreover, this paper extends to a different supply chain with a reverse channel leadership and a different product return policy for testing the robustness.

Several interesting and important results are derived in this paper. Firstly, it is found that the relative pricing are largely relied on the costs of two channels. Secondly, the platform supply chain may benefit from a pure channel rather than the dual-channel when this channel enjoys a relatively low cost and/or a sufficiently high consumer preference. Then, the platform and the supplier act contradictorily when selecting their optimal online channel modes. To be specific, the platform motivates to choose the online reselling mode when both the commission rate and the slotting fee are relatively low, whereas the supplier is likely to select the online agency mode under this circumstance. Finally, a win-win situation in regards to the optimal online channel mode for two parties is achievable with numerical experiments.

Based on the analytical studies, the results derived in the authors’ work can provide managerial insights to assist the supplier and the platform company in determining the operational decision and selecting the optimal online channel mode to deal with consumer returns. In addition, appropriate commission rate along with slotting fee will make both parties achieve a win-win situation in determining their optimal online channel mode.

To the authors’ best knowledge, this paper makes the first move to determine the optimal online channel mode in the content of consumer returns and study how it is affected by different product return policies.

This study aims to elucidate the machining properties of low-cost expanded clay-reinforced syntactic foams by using different neural network models for the first time in the literature. The main goal of this endeavor is to create a casting machining-neural network modeling flow-line for real-time foam manufacturing in the industry.

Samples were manufactured via an industry-based die-casting technology. For the slot milling tests performed with different cutting speeds, depth of cut and lubrication conditions, a 3-axis computer numerical control (CNC) machine was used and the force data were collected through a digital dynamometer. These signals were used as input parameters in neural network modelings.

Among the algorithms, the scaled-conjugated-gradient (SCG) methodology was the weakest average results, whereas the Levenberg–Marquard (LM) approach was highly successful in foreseeing the cutting forces. As for the input variables, an increase in the depth of cut entailed the cutting forces, and this circumstance was more obvious at the higher cutting speeds.

The effect of milling parameters on the cutting forces of low-cost clay-filled metallic syntactics was examined, and the correct detection of these impacts is considerably prominent in this paper. On the other side, tool life and wear analyses can be studied in future investigations.

It was indicated that the milling forces of the clay-added AA7075 syntactic foams, depending on the cutting parameters, can be anticipated through artificial neural network modeling.

It is hoped that analyzing the influence of the cutting parameters using neural network models on the slot milling forces of metallic syntactic foams (MSFs) will be notably useful for research and development (R&D) researchers and design engineers.

This work is the first investigation that focuses on the estimation of slot milling forces of the expanded clay-added AA7075 syntactic foams by using different artificial neural network modeling approaches.

In this paper, an improved multiobjective particle swarm optimization (PSO) algorithm is proposed to optimize a three-phase, 12-slot, 19-pole, yokeless axial-field flux-switching permanent magnet (YASA-AFFSPM) motor.

Based on the structural characteristics of the YASA-AFFSPM, a mathematical model is established to calculate the main size of the YASA-AFFSPM motor. The split ratio, stator axial length, sandwiching pole angle, rotor pole angle, PM arc and number of conductors per slot are selected as optimization variables. Also, the efficiency, power factor, cogging torque and average torque are considered as the optimization objectives. The objectives are optimized by combining the improved multiobjective PSO algorithm with electromagnetic calculation.

Based on the proposed algorithm, the investigated motor is optimized. The on-load efficiency, power factor and average torque of the motor performance have increased by 0.87%, 3.13% and 10.39%, respectively. Moreover, the cogging torque and slot fill factor have undergone decreases of 8.57% and 3.34%, respectively. Finally, the effectiveness of the algorithm is verified using experiment results.

So far, no comprehensive report has been observed on the optimization of the YASA-AFFSPM motor using evolutionary algorithms and the study of the effect of the motor parameters. Therefore, in this paper, the authors decided to investigate the effect of YASA-AFFSPM motor parameters and improve motor performance with the improved PSO method.