In digital manufacturing, data-driven methods are promising to revolutionize various decision-making processes. However, the relationships between variables in high-dimensional data of data-driven decision-making methods are only correlations. Important causal relationships and knowledge between process variables are not considered. Therefore, existing data-driven systems are unstable, which could result in unreliable and dangerous decisions. To establish a stable decision-making model for complex processes with high-dimensional data, a causal-based decision-making framework that combined causal relationships and knowledge between key manufacturing variables was proposed. The causal relationships between state, decision, and objective data were established in the form of a direct acyclic graph formed by breaking an unexcepted loop between variables using a shadow objective variable. Then, causal knowledge of high-dimensional states was introduced to the neural network, forming a stable decision-making model. Compared with data-driven methods used in robotics and manufacturing scenarios, the proposed framework provided better and more stable decisions, particularly in noised environments.

This study focuses on the simultaneous estimation of unknown inputs (UIs) and states of nonlinear discrete-time heterogeneous multi-agent system with homologous UIs. Based on unscented Kalman filter (UKF), a minimum-variance unbiased filter for the UIs and state estimation is proposed for the multi-agent system with homologous UIs. Compared with previous studies, this paper proposes a general model that is applicable in practice. The neighbors' information of the homologous UIs are utilized to construct sigma points of the homologous UIs. In the case study, a target tracking problem is used to verify the advantage of the UKF-based distributed filter.

The state prediction of key components in manufacturing systems tends to be risk-sensitive tasks, where prediction accuracy and stability are the two key indicators. The physics-informed neural networks (PINNs), which integrate the advantages of both data-driven models and physics models, are deemed as an effective approach and research trends for stable prediction; however, the potential advantages of PINN are limited for the situations with inaccurate physics models or noisy data, where the balancing of the weights of the data-driven model and physics model is very important for improving the performance of PINN, and it is also a challenge urgently to be addressed. This article proposed a kind of PINN with weighted losses (PNNN-WLs) by uncertainty evaluation for accurate and stable prediction of manufacturing systems, where a novel weight allocation strategy based on uncertainty evaluation by quantifying the variance of prediction errors is proposed, and an improved PINN framework is established for accurate and stable prediction. The proposed approach is verified with open datasets on tool wear prediction, and experimental results show that the prediction accuracy and stability could be obviously improved over existing methods.

This study addresses the problem of the estimation of state when heterogeneous multiagent systems are affected by homologous unknown inputs (UIs). Homologous UIs refer to identical UIs affecting different agents. An improved semidistributed filter based on previous research is proposed. The improved filter uses neighbors’ information for UI estimation but not state estimation. A necessary and sufficient condition for the proposed filter to achieve minimum-variance unbiased estimation is presented and proven. Moreover, the asymptotic stability of the filter is analyzed. A sufficient condition of the asymptotic stability is presented and proven. The theoretical and numerical analyses indicate that the proposed filter has less communication pressure, fewer calculation requirements, and better estimation performance compared with the existing solutions.Note to Practitioners—In the industry, homologous unknown inputs (UIs) exist in many different systems. For example, the same ambient temperature affects the performance of every battery in a battery pack. Similarly, the same wind power can affect different aircrafts flying in the same region. Temperature and wind power can be considered the homologous UIs of a multiagent system. Estimation of homologous UIs is important because of the latter’s massive impact on the system. In this study, data transmission delay and packet loss are ignored. Hence, the study is limited to low-rate systems. Moreover, nonlinear filters must be studied further in future work.

Cross conditions prediction is a prevalent problem in manufacturing area, where tool wear prediction is a typical one. Existing data-driven methods for tool wear prediction mainly focus on cutting conditions with small variations, which encounters much difficulty under cross conditions with large variations, and the essential is the difference of both marginal distribution and conditional distribution of the data under cross conditions. To address this issue, this article proposes a meta-invariant feature space (MIFS) learning method, where invariant feature space is constructed for paired tasks to close marginal distribution, whose nature law under cross conditions is learned by meta-learning, i.e., MIFS, which can be adapted to achieve accurate tool wear prediction under cross conditions with a small number of new samples. Experimental results provided positive confirmation on the feasibility and accuracy of the proposed method, which can also be readily extended to regression and classification problems in other fields.

Complex structural parts are pervasive and playing an important role in the aircraft manufacturing area. In order to improve the machining efficiency, the cutting parameter optimization of complex structural parts during the machining has always been a problem in manufacturing industry. At present, the cutting parameters are usually optimized based on the final state of complex structural parts and remain unchanged during the machining process, which may not consider the cutting parameter optimization of workpiece in the intermediate machining process. Thus, a cutting parameter optimization method based on dynamic machining features for complex structural parts is proposed to improve the machining efficiency and guarantee the product quality during the machining process. The interim geometric state of each machining occasion is constructed in order to analyze the chatter stability. Then, the cutting parameters are optimized using a genetic algorithm within the limits of chatter stability.

The production of aircraft structure parts is featured by multiple varieties and small batches, which imposes significant challenges for the representation of process knowledge. Feature based method is an effective way as the process knowledge carrier. Although the parts are different from each other, they are composed of similar geometric features with similar machining processes. This paper introduces the concept of "dynamic machining feature" which is formed in the machining process and influenced by various real operations. In this paper, the process knowledge and interim geometric information are associated based on dynamic machining features. An ontology-based method has been adopted to represent relevant information of dynamic machining features. The proposed approach can speed up process decision and facilitate process optimization.

Small batch and multiple variety production mode and changing machining conditions call for dynamic inspection to guarantee machining quality. Dynamic inspection by considering real time monitoring information is a promising approach, but there is no available technology for integrating real time monitoring and inspection. To address this issue, this paper proposes a method of integrating monitoring and inspection based on intelligent software agents and manufacturing features. An agent-based approach is applied to develop an integrated framework, while manufacturing features are used as the information carrier to represent and connect monitoring and inspection information. Dynamic inspection is triggered according to the analysis results of real time monitoring. A prototype system has been developed to implement and validate the proposed method.

Manufacturing of aircraft structural parts has the characteristics of multi-varieties, complex structure, and small batch which result in high difficulty and workload of the processing planning. When the project leader is deciding the assignment of processing planning tasks, two main factors which affect the reliability of the processing plan need to be considered: the capability of the planners and the complexity of the structural parts. This paper proposes a fuzzy comprehensive evaluation method to evaluate these two factors. The reliability of processing plans is predicted by using the two impact factors and evaluated by using the machining results both based on the fuzzy comprehensive method. A project case study is taken to demonstrate the usability of the proposed method and the test results show that the proposed prediction method is feasible and practicable.

Due to the lack of efficiency communication and collaboration among each manufacturing department, it makes long lead-time and high cost for the manufacturing of aircraft structure parts which have the characteristics of multi-varieties, complexity structure, and small batch. In order to address these challenge issues, a feature-based and multi-agents-based manufacturing collaborative framework which includes process planning agent, NC programming agent, fixture designing agent, cost estimation agent and production management agent for aircraft structural parts is proposed, and the collaboration among each other is realized based on feature, the collaboration between fixture design and NC programming is described in detail. The prototype developed based on the framework has been utilized in a large aviation enterprise, and the result shows that the lead-time and cost are decreased.

NC machining time of a part depends on its geometry and process plan, NC program, and machine characteristics. Since the integration of geometry and process plan, NC program, and machine characteristics is difficult during NC machining time forecasting, existing commercial software tools and research systems do not fully consider these factors, and therefore the machining time forecasting accuracy is low. In order to address this challenging issue, this paper proposes a feature-based model for NC machining time estimation. Experiment results shown that the proposed approach is feasible and practical.

Because of the characteristics of large dimensions, complex structure and high machining accuracy in aircraft structural parts, iterations frequently happen in the parts design process, which increase the R&D time and costs substantially. In response to the situation, a cooperative design framework based on multi-agent for aircraft is proposed in this paper. Agent of detailed design, feature recognition, manufacturability analysis and etc. are included in the framework. According to the constraints upstream, new feature or structure will be constructed to confirm a new design scheme during the design process. Through the local feature recognition, machining features are recognized from the design feature model simultaneously, and then employed to analyze manufacturability and evaluate costs, in order to obtain an optimal design scheme. An original system has been developed and partially tested in an aviation enterprise. The result shows that the methodology is practicable.

Multiple amplitude shaped offset quadrature phase-shift keying (MA-SOQPSK) modulation based on multiple nonlinear power amplifiers (NLPAs) provides both high power-added efficiency and spectrum efficiency. However, the phase misalignment between the different branches of MA-SOQPSK signal in the process of the modulation can significantly degrade the bit-error-rate (BER) performance. A novel scheme of correcting this phase error is proposed in this paper, which contains a phase lock loop (PLL) in the modulation terminal. Theoretical estimation and simulation on the variance of the phase error are also given to validate our method. The BER simulation result shows that our scheme improves the performance more than 3 dB at 10"3, and the improvement becomes more distinct as the SNR.

Remote design and manufacturing is one of the enabling technologies for implementation of virtual enterprise. Remote collaborative design and verification of NC machining process plays an important role in remote design and manufacturing. In this paper the authors present a Web-based architecture for remote collaborative design of NC machining process and NC program, a virtual machining and measuring cell is introduced and developed. The key technology for the virtual machining and measuring cell development is discussed, including digital workpiece description, cutting force prediction, etc. The feasibility has been verified by case studies.