Fault root cause location is crucial role for industrial system safe operation. However, most industrial processes are time-varying due to operating mode changes, demand transitions, and equipment degradation, causing system causality evolution. These variations can reduce the performance of conventional fault root cause location methods. To address this issue, a spatial–temporal adaptive causality graph generation method (STACGG) is proposed to update the causality graph for fault root cause location. In the STACGG method, an edge aggregation masking operator is designed, in which the common part of the causality can be well inherited, while the customized part is online learned within the allowable maximum causality variation range, thus achieving steady adaptive progressive causality graph update. First, the most similar historical causality graphs are matched by contrasting the temporal and spatial characteristics of the pairwise node feature. Then, an edge aggregation-based graph generator (EAGG) is developed to identify a compact edge structure between the edge intersection and the edge union of these similar causality graphs. Enabling the edge intersection as the learning lower bound means inheriting common part of causality while taking the edge union as the learning upper bound represents steadily learn the customized part within the maximum causality variation range. To achieve it, the EAGG is formulated as an optimization task that maximizes the mutual information entropy between a GNN’s fault detection and the possible edge structure distribution. Then, to enhance the expression power and the interpretability of the causality graph, the posterior data information and the prior physical knowledge are combined as the inductive bias and the learning bias to fine-tune the causality graph for graph performance boosting. Finally, the fault root cause location performance is validated on two real-word industrial cases.

At present, working conditions are becoming more complex, and operation policy requirements are more diverse in process system engineering. To control a process problem, a balance must be found between speed and stability, in that operations should sometimes be faster and other times smoother. Traditional controllers, such as PID and model predictive control are applied in various problems, and some parameters in controllers can be used to represent the operation policy. However, there can be difficulties in tuning parameters, and time costs of online calculation. This article proposes parametric deep reinforcement learning (PDRL) to replace traditional controllers. PDRL has two parts. A vanilla DRL framework is adapted to solve the setpoint tracking problem. With a state and a reward function and robust training tricks, trained agents can be applied to more generalized working conditions. Base agents of different operation policies are trained in advance. With target performance from operators, the target policy can be fitted by base agents with a set of weights, which are first optimized by minimizing the squared error between the target and fitted policy in a basic task, and applied to generalized conditions. A shell benchmark problem is chosen as a case study, whose results show that PDRL has feasibility and stability both in basic and generalized tasks, even in a noisy environment.

At an urban intersection, traffic performance is highly influenced not only by signal timing schemes but also by lane assignments. However, dynamic adjustment of lane configurations remains understudied due to the prevalent use of predetermined lane arrangements. To address this research gap, we propose the TransLane framework, which collaboratively optimizes time and space resources within an intersection by implementing dynamic lanes. We model each intersection as an agent and apply a hierarchical recommendation approach to these agents. Additionally, the framework addresses the challenge of training an agent in the absence of sufficient data by using transfer learning techniques. Specifically, a new intersection agent can be trained from pre-trained agents via model and sample transfer mechanisms, which facilitate the reuse of traffic control knowledge from similar intersections. The efficacy and superiority of TransLane are demonstrated through simulation studies that show it outperforms models that optimize space or time resources independently using widely adopted traffic control optimization methods.

In multiunit industrial plant-wide processes, how to simultaneously model the dynamic correlations within units and the graph-structured interactions between units has not been explored by existing distributed monitoring methods. To address this issue, this work proposes a novel hybrid-order graph embedded distributed encoder-decoder (HGDED) for plant-wide process monitoring. Firstly, the whole process is decomposed into multiple operation units and characterized as a directed graph based on process knowledge. Then, distributed node encoder and decoder are developed under the sequence-to-sequence framework to capture the intra-unit temporal dependence. Between the encoder and decoder, a novel hybrid-order graph convolutional network (HGCN) is designed to simultaneously embed the inter-unit spatial dependence. Through multireceptive field graph convolution and spatial recurrent updates, HGCN can effectively exploit hybrid-order neighbor information to capture multi-cascaded interactions between units. Moreover, a spatial order optimizer is innovatively proposed to automatically learn the optimal discrete orders for different nodes, which helps HGCN better capture significant features from higher-order neighbor nodes. With concurrent analysis of temporal and spatial dependencies, the representation learning ability of HGDED is enhanced, thereby improving the monitoring performance. Finally, the effectiveness of the proposed method is demonstrated through the Tennessee Eastman process and a real-world air separation process.Note to Practitioners—The complex process characteristics and topology structure have imposed considerable challenges on monitoring multiunit industrial plant-wide processes. In practice, it is necessary to extract both the intra-unit temporal dependence and the inter-unit spatial dependence for distributed monitoring. Existing studies fail to consider these two important characteristics at the same time, which may lead to information loss and performance degradation. In this work, the proposed HGDED not only develops distributed node encoder and decoder to model the dynamic correlations within units, but also incorporates a HGCN to capture the interactions between units. In particular, HGDED explicitly models the whole process into a directed graph, which can characterize structural relationships in a more reasonable and fine-grained way. Considering the multi-cascaded information transmission between units, the HGCN is designed to exploit hybrid-order neighbor information for graph embedding, in which the optimal discrete orders are automatically determined by the spatial order optimizer. As a consequence, HGDED can better model process behaviors and extract more representative features to improve monitoring performance. In addition, HGDED follows an end-to-end training fashion, which is convenient for engineers to implement. The proposed method has been validated on two industrial cases, and it is suitable for monitoring various practical industrial plant-wide processes.

Causality in an industrial process provides insights into how various process variables interact and affect each other within the system. It reveals the underlying mechanisms of industrial processes, which ensures predictive reliability and facilitates physical interpretability. However, existing causality-based techniques have limitations, as they neglect the temporal factor in causal description, introduce spurious causal associations in causal discovery, and fail to consider the spatial-temporal synchronicity in causal utilization. To address these issues, this article proposes a spatial-temporal causality modeling approach. A novel spatial-temporal causal digraph (STCG) is proposed to describe causal dependencies among process variables, which considers both spatial and temporal factors encompassing causal relationships and time delays. The STCG identification procedure is formulated as a Markov decision process, and knowledge-data guided reinforcement learning is developed to acquire the optimal identification policy and avoid spurious causal associations. With the identified STCG, a graph attention gate recurrent unit (GAGRU) is constructed for spatial-temporal process modeling, which is able to capture the synchronous evolution of industrial data in spatial-temporal dimensions. Finally, the effectiveness of the proposed modeling approach is verified by applying to two real industrial cases, including soft sensing for a sulfur recovery unit and anomaly detection for an argon distillation system. The experimental results demonstrate that the STCG-based industrial process modeling outperforms classical and state-of-the-art comparison methods in terms of reliability and interpretability.

Sensor faults are non-negligible issues for soft sensor modeling. However, existing deep learning-based soft sensors are fragile and sensitive when considering sensor faults. To improve the robustness against sensor faults, this article proposes a deep subdomain learning adaptation network (DSLAN) to develop a sensor fault-tolerant soft sensor, which is capable of handling both sensor degradation and sensor failure simultaneously. Primarily, domain adaptation works for process data with sensor degradation in industrial processes. Being founded on the basic structure of deep domain adaptation, a novel subdomain learner is added to automatically learn the subdomain division, enabling DSLAN adaptable to multimode industrial processes. Notably, the subdomain structure of each sample follows a categorical distribution parameterized by output of the subdomain learner. Based on the designed subdomain learner, a new probabilistic local maximum mean discrepancy (PLMMD) is presented to measure the difference in distribution between source and target features. In addition, a generator for failure data imputation is integrated in the framework, making DSLAN handle sensor failure simultaneously. Finally, the Tennessee Eastman (TE) benchmark process and two real industrial processes are used to verify the effectiveness of the proposed method. With the fault tolerance ability, soft sensing technology will take a step toward practical applications.

In this article, a novel approach to hierarchical implementation of output feedback robust model predictive control is proposed for the linear polytopic uncertain model. One optimization problem for minimizing the performance index is followed with the other assessing estimation error set (EES). The two problems are posed in a lexicographic order. Since in the latter problem, the controller parametric matrices are retaken as the degrees of freedom for the optimization, a much less conservative EES is calculated. Therefore, by applying the new approach, the control performance can be greatly improved as compared with the earlier schemes without lexicographic optimization. The proposed approach is proven to be recursively feasible, and the closed-loop stability is specified by the notion of quadratic boundedness. The result is verified through two numerical examples.

Optimal iterative learning control (OILC) has been recognized as an excellent model-based means for regulating batch process with abundant successful applications reported in the past decades but also received considerable criticisms for its poor robustness against model mismatch that is common for many industrial situations. Despite numerous attempts to address the issue, many of them are still not able to yield satisfactory control performance particularly in the presence of a possible combination of time-varying uncertainties and conservatively designed controllers, which may compromise the learning mechanism, hence rendering the robustness issue of OILC far from well explored. This article intends to investigate the aforementioned issue by proposing a new OILC method resting upon the minimization of a dynamic upper bound on tracking error which is distilled from better exploitation of the time variation of uncertainties. We also show that the problem can be formulated in the framework of convex–concave game that can be efficiently solved by a subgradient method with an excellent balance of optimality and computation time. Such a formulation enables us to gain: 1) guaranteed monotonic convergence on tracking error; 2) remarkably reduced conservatism on controller synthesis; and 3) controllable computation complexity. It is further shown that the proposed method is capable of handling nonlinearity, for example Volterra system, a classic representation of nonlinear process. The efficacy of the method is verified by numerical experiments on a continuous stirred tank reactor model.

A model predictive controller was designed in this study for a single supply chain unit. A demand model was described using an autoregressive integrated moving average (ARIMA) model, one that is identified on-line to forecast the future demand. Feedback was used to modify the demand prediction, and profit was chosen as the control objective. To imitate reality, the purchase price was assumed to be a piecewise linear form, whereby the control objective became a nonlinear problem. In addition, a genetic algorithm was introduced to solve the problem. Constraints were put on the predictive inventory to control the inventory fluctuation, that is, the bullwhip effect was controllable. The model predictive control (MPC) method was compared with the order-up-to-level (OUL) method in simulations. The results revealed that using the MPC method can result in more profit and make the bullwhip effect controllable.