Inner-phase and inter-phase analysis based operating performance assessment and nonoptimal cause identification for multiphase batch processes

Due to environmental fluctuations, the operating performance of complex industrial processes may deteriorate and affect economic benefits. In order to obtain maximal economic benefits, operating performance assessment is a novel focus. Therefore, this paper proposes a whole framework from operating performance assessment to nonoptimal cause identification based on partial-least-squares-based Granger causality analysis (PLS-GC) and Bayesian networks (BNs). The proposed method has three main contributions. First, a multiblock operating performance assessment model is established to correspondingly extract economic-related information and dynamic information. Then, a Bayesian network structure is established by PLS-GC that excludes the strong coupling of variables and simplifies the network structure. Lastly, nonoptimal root cause and and nonoptimal transmission path are identified by Bayesian inference. The effectiveness of the proposed method was verified on Benchmark Simulation Model 1.

In order to realize the operating performance assessment of multi-mode industrial production process with both qualitative and quantitative information, an online operating performance assessment method based on Rough Set (RS) is proposed. Operating performance assessment consists of offline modeling and online assessment. Firstly, the modeling data of different modes are collected to establish the mode rule base, and operating performance assessment rule base based on RS respectively; Then, the online operating performance assessment is implemented by mode identification of online sample data and rule matching; When the process is operating at a non-optimal performance, non-optimal cause identification is carried out through calculating the variable difference degree between the current non-optimal data and the rule in the optimal rule base, which offers guidance to the operators to bring the process back to the optimal performance.

To improve the performance of process monitoring method based on support vector data description (SVDD), a local kernel distance-SVDD (LKD-SVDD)-based process monitoring method for multiphase batch processes is proposed. First, the batches are partitioned into phases by using a warped k-means (WKM)-based clustering approach to addresses the sequential characteristics of batch process data. Then, considering the nonlinear and high-dimensional characteristics of batch process data, an LKD-SVDD-based modeling method is proposed for online monitoring and fault diagnosis, which effectively improves the computation efficiency and the process monitoring performance. A fed-batch penicillin fermentation process is used to demonstrate the feasibility of the proposed method.

Operating optimality assessment and nonoptimal cause identification for non-Gaussian multimode processes with transitions

The Penataran Train is the train that has the highest passenger volume among the Local Economy Trains in East Java Province. The Penataran train is very much needed by people, but if the operational and management is not handled properly, it may affect on the effectiveness and efficiency of the transportation system as a whole. Operational performance assessment is needed to improve the services to users of the Penataran Train in the future. Operational performance was assessed on several aspects, including load factor, travel time, frequency, headway, and ticket prices. The purpose of this study was to analyze the operational performance and prices of the Penataran Train on Surabaya City – Blitar route. The results showed that the average load factor was 61% and met the standard of SE No. 24 of 2020. The average rate of the train arrival time of 99% was in accordance with the SIPOKA standard. In contrast, the average rate of departure time of 83% was not in accordance with the SIPOKA standard. The frequency of 8 trips per day with a headway of 198.75-202.25 minutes was not meet the standards in KM No. 22 of 2003. The point of intersection between the respondents' ATP and WTP was at a price following the ticket price stipulated in PM No. 21 of 2020, which was Rp10,001–Rp15,000

Modeling basic freeway section level-of-service based on travel time and reliability

This chapter looks at performance assessment in the context of the day-to-day functioning of irrigation and drainage systems, and shows how performance assessment can be integrated into the management processes of irrigation and drainage systems. An explanation of service delivery in irrigation and drainage is provided, followed by a brief discussion of the impact that different formulations of the physical infrastructure and management structure can have on service delivery. Approaches are then formulated for operational and strategic performance assessment of service delivery in these different contexts. A key focus is on performance-oriented management, the basic components of which are: (i) the specification of the services and the level of service provision by the irrigation service provider to the water users; (ii) agreement between the water users and the irrigation service provider on the rights and responsibilities of the water users, particularly in relation to payment for services received; (iii) procedures for monitoring the services provided and responsibilities fulfilled; and (iv) procedures for evaluating the services provided and the responsibilities fulfilled.

This study is aim to analyze of Operational Performance Measurement using Balanced Scorecard method. Subject of this research is RSBT Pangkalpinang. This balanced scorecard is a tool that showed the hospital performance based on four perspectives namely financial, customer, internal business process, and learning and growth perspective. This research uses a quantitative-descriptive method. This study using primary and secondary data from 2018-2020. Primary data collection techniques through the distribution of questionnaires and interviews with respondents. 
 The results of the analysis prove that the overall operational performance assessment at the RSBT Pangkalpinang uses a balanced score card that is satisfying and has a good performance mesurement.

Operating performance assessment of industrial process based on MIC-graph convolutional networks with local slow feature analysis.

In modern hot rolling process, operating performance assessment is of great practical significance for guiding the production adjustment for operators. From the perspective of classification, operating performance assessment is a multi-class classification problem. Since support vector data description (SVDD) is a one-class classifier, conventional methods usually construct an independent SVDD model for each class, which ignores the correlation among different classes. The hyperspheres of different classes may not be isolated but overlapped. If a test sample exists in the overlapping region, how to determine which class it belongs to is a knotty problem. Moreover, conventional methods treat all samples equally, but in practice, the sample number of different classes can be imbalanced, which will affect the classification performance of SVDD. In this study, an operating performance assessment method based on weighted multi-sphere SVDD (WMSVDD) is proposed for solving the aforementioned issues. WMSVDD considers the interactions among different classes in a unified way, optimizes the hyperspheres of different classes globally, and introduces a weight coefficient to the model for eliminating the affects of uneven class sizes. Simulation results on a real hot rolling process illustrate the effectiveness of the proposed method comparing to the traditional multi-class SVDD.

Purpose: Persistent underrepresentation of racially, ethnically, and socioeconomically diverse populations, coupled with protracted enrollment timelines, continues to compromise the generalizability, external validity, and equity of clinical research. Blood sampling, among the most frequently mandated in-person procedures, presents logistical barriers that inadvertently impede participation, retention, and biospecimen quality. Travel requirements, scheduling constraints, and inconsistent specimen handling disproportionately burden older adults, rural populations, and participants with limited temporal or transportation flexibility. We describe a novel, AI-integrated patient recruitment paradigm operationally coupled with decentralized clinical research phlebotomy and supported by digital chain-of-custody documentation and quality oversight. We present key operational outcomes associated with this integrated approach. Design: We conducted a comprehensive operational evaluation of an AI-enabled recruitment workflow integrated with mobile and community-based clinical research phlebotomy across multiple trial sites. Outcomes were compared descriptively against historical controls utilizing conventional, site-centric recruitment and phlebotomy workflows, facilitating assessment of enrollment efficiency, demographic representativeness, and operational performance. Result: The AI-enabled recruitment system identified 2,350 potential candidates and facilitated enrollment of 320 participants, representing 64% of the total study population (N = 500). Compared with historical controls, time to complete enrollment decreased from 12 months to 8 months 33% reduction. Concurrently, the proportion of enrolled participants from underrepresented populations increased from 22% to 38%, demonstrating substantial enhancement in trial representativeness without compromising operational feasibility. An AI-integrated recruitment strategy coupled with decentralized clinical research phlebotomy can substantially accelerate enrollment while enhancing demographic representativeness. When combined with standardized digital workflows and rigorous chain-of-custody oversight, this model preserves biospecimen integrity and data reliability, positioning phlebotomy as a strategic enabler of equitable, efficient, and high-quality clinical research.

Dual learning-based online ensemble regression approach for adaptive soft sensor modeling of nonlinear time-varying processes

We propose a new batch bioprocess monitoring approach which combines the Gaussian mixture model (GMM) with the multiway analysis in localized discriminant embedding subspace for detecting and classifying different types of faults. Due to the inherent process multimodality within abnormal points, the capability of traditional multimay locality preserving projection (MLPP) is degraded. For the routine bioprocess operation with abnormal events, GMM is used to partition the training set with various types of faults into different clusters. Three localized distance matrices reflecting the relationships of nearby points are then computed. The extracted leading discriminant embedding directions can not only separate the normal and faulty data by maximizing the distance among nearby data points from different clusters, but also preserve the intrinsic geometrical structure within the multiple faulty clusters by minimizing the small distance and penalizing the large distance among neighboring data from the same cluster. The newly developed multiway localized discriminant embedding (MLDE) approach is applied to two test scenarios in the fed-batch penicillin fermentation process and compared with the conventional MLPP method. The results demonstrate that the MLDE approach is superior to the MFDA method in detecting abnormal events and classifying different types of faults in fed-batch process with higher accuracy and sensitivity.

We consider a hybrid modelling approach based on the combination of prior knowledge, in the form of mechanistic models, with tools for the extraction of knowledge from operating data: the first component captures first-principles system behavior features, while the second one accounts for the differences found between mechanistic predictions and real data. The empirical modelling methods tested include backpropagation Artificial Neural Networks (ANN), Multivariate Adaptive Regressive Splines (MARS) and Regression Analysis (RA). In the proposed hybrid structure, mechanistic predictions provide additional inputs for the empirical module. To evaluate the performance of the various combinations of modules and hybrid approaches, we consider two simulated case studies (involving a CSTR with a reversible reaction and a fed-batch penicillin fermentation process). In all of the above, our hybrid structure based on a mechanistic module together with an empirical component has outperformed other approaches, and the mechanistic model/MARS combination resulted in lowest overall prediction errors.

Hybrid modeling through the combination of mechanistic models and MARS