MantaRay-ProM: An efficient process model discovery algorithm

This research study aims to introduce chaos theory into the Manta Ray Foraging Optimization (MRFO) Algorithm and optimize a real-world design problem through the chaos-enhanced versions of this method. Manta Ray Foraging Optimization algorithm is a bio-inspired swarm intelligence-based metaheuristic algorithm simulating the distinctive food search behaviors of the manta rays. However, MRFO suffers from some intrinsic algorithmic inefficiencies such as slow and premature convergence and unexpected entrapment to the local optimum points in the search domain like most of the metaheuristic algorithms in the literature. Recently, random numbers generated by chaos theory have been incorporated into the metaheuristic algorithms to solve these problems. More than twenty chaotic maps are applied to the base algorithm and ten best performing methods are considered for performance evaluation on high-dimensional optimization test problems. Forty test problems comprising unimodal and multimodal functions have been solved by chaotic variants of MRFO and extensive statistical analysis is performed. Furthermore, thermo-economic design optimization of an air-fin cooler is maintained by the chaotic MRFO variants to assess their optimization capabilities over complex engineering design problems. Ten decisive design variables of an air fin cooler are optimized in terms of total annual cost rates and optimum solutions obtained by five best chaotic MRFO algorithms are compared to the preliminary design. A significant improvement is observed in the objective function values when MRFO with chaotic operators is applied to this considered thermal design problem.

The decarburization of heat and transport using electric vehicles (EVs) and heat pumps (HPs) needs an important investment in low-voltage (LV) networks. This article proposes an optimization approach to estimate the availability of headroom for domestic EV charging and optimization on LV networks under various penetrations of heat pumps. The proposed optimization approach is called dynamic differential annealed optimization (DDAO). The objective of the proposed approach is to maximize the hosting capability of low-voltage networks for EVs and HPs by optimizing the flexibility of EV charging. The DDAO is used to optimize different aspects of EV charging and LV networks. The DDAO approach is implemented in three steps: (a) HP headroom assessment, (b) EV optimization, and (c) validation. The performance of the proposed method is validated on a MATLAB platform and compared with the existing Binary Spring Search Algorithm (BSSA), Manta Ray Foraging Optimization Algorithm (MRFO), and Chaos Game Optimization methods. In comparison to the total prices of 3.5£ in BSSA, 4.5£ in MRFO, and 5.5£ in CGO, the proposed DDAO method achieves a lower total price of 3.2£. This suggests that the DDAO method is more cost-effective than the existing techniques.

Virtualization is a transformative technology that enables multiple virtual machines to operate concurrently on a single physical host, thereby rationalizing resource utilization in cloud data centers. These data centers consist of multiple servers and consume a significant amount of energy, which requires cloud providers to have mechanisms that optimize the placement of virtual machines in these servers. An optimal placement allows for a decrease in the energy consumed and an increase in the resource usage rate. In this paper, we introduce a novel algorithm based on a recent bio-inspired metaheuristic called Manta Ray Foraging Optimization (MRFO) to solve the virtual machine placement problem. Although MRFO was recently proposed in order to solve optimization problems for engineering applications, to the best of our knowledge, it has never been used to solve the virtual machines placement problem in the cloud. Accordingly, this paper investigates for the first time an enhanced version of the MRFO algorithm to solve the virtual machine placement problem. The proposed algorithm is evaluated with CloudSim toolkit under various performance metrics, including energy consumption, resource utilization, and number of active servers. Simulation results show that our approach outperforms five well-known metaheuristics widely adopted in this field, namely Simulated Annealing (SA), Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Salp Swarm Algorithm (SSA), and Whale Optimization Algorithm (WOA).

Flexible business processes can often be modelled more easily using a declarative rather than a procedural modelling approach. Process mining aims at automating the discovery of business process models. Existing declarative process mining approaches either suffer from performance issues with real-life event logs or limit their expressiveness to a specific set of constaint types. Lately, RelationalXES, a relational database architecture for storing event log data, has been introduced. In this paper, we introduce a mining approach that directly works on relational event data by querying the log with conventional SQL. By leveraging database performance technology, the mining procedure is fast without limiting itself to detecting certain control-flow constraints. Queries can be customised and cover process perspectives beyond control flow, e.g., organisational aspects. We evaluated the performance and the capabilities of our approach with regard to several real-life event logs.

The identification of key information hidden in non-stationary signals is challenging in various fields such as logistics and transportation, biomedicine, and fault diagnosis. To facilitate this identification, we propose a back propagation neural network (BPNN) classification and recognition algorithm based on wavelet threshold denoising (WTD) and manta ray foraging optimization (MRFO) algorithm for the first time. The algorithm first performs WTD on the original signals to obtain denoised signals. Subsequently, the MRFO algorithm is utilized to optimize the initial weights and thresholds of the BPNN. On the base of this, the optimization model is finally obtained to classify and recognize the key information in the non-stationary signals. The comparative experimental results indicate that the proposed WTD-MRFO-BPNN algorithm can be utilized to availably recognize the key information hidden in non-stationary signals. The recognition accuracy reaches 97.25%.

In this paper, an accurate model of proton exchange membrane fuel cell (PEMFC) for optimal identification of PEMFC parameters has been developed. The optimization methodology is based on the modified version of Manta Ray Foraging Optimization (MMRFO) technique for minimizing the sum of squared errors (SSE) between the Experimentally measured stack voltage and the estimated voltage produced by the optimized model. In the modified methodology, the sine-cosine method has been utilized to enhance the global searching capability in the exploration phase and the local searching capability in the exploitation phase of the MRFO algorithm. In order to validate the effectiveness of the suggested methodology, four different case studies comprising standard benchmark 250 W PEMFC, BCS-500 W PEMFC, SR-12 500 W FC, and 1 kW Temasek stacks were utilized, and the attainments have been compared with the measured polarization characteristics. The attainments have been intensively compared with several metaheuristic algorithms (MA) including Tree growth Algorithm (TGA), Grey wolf optimizer (GWO), Whale optimization algorithm (WOA), Salp swarm algorithm (SSA), and original Manta Ray Foraging Optimization (MRFO), to confirm the superiority of the MMRFO against the compared techniques. The obtained results give a satisfactory agreement between the MMRFO-based model and the experimentally measured data. Finally, the achievements confirmed the effectiveness of the MMRFO over the basic MRFO algorithm and other novel metaheuristic algorithms in identifying PEMFC parameters.

This review paper delves into the original MRFO and its variants, focusing on single-objective algorithms including but not limited to hybrid algorithms, learning strategies, multiple populations and dynamic parameter adjustment, highlighting the improvements made to enhance the algorithm's efficiency in global optimization, accelerate convergence rates, and improve its capacity to evade local optima. MRFO has emerged as an effective tool for solving complex optimization problems across various domains, including energy optimization, biomedical field, engineering problems, and others. A comprehensive analysis of applications of MRFO in different fields is provided, emphasizing its adaptability and efficacy. The paper concludes with a discussion on the challenges faced by MRFO and potential future research directions, aiming to consolidate the current research status and guide future investigations.

Process mining is the area of research that embraces the automated discovery, conformance checking and enhancement of process models. Declarative process mining approaches offer capabilities to automatically discover models of flexible processes from event logs. However, they often suffer from performance issues with real-life event logs, especially when constraints to be discovered go beyond a standard repertoire of templates. By leveraging relational database performance technology, a new approach based on SQL querying has been recently introduced, to improve performance though still keeping the nature of discovered constraints customisable. In this paper, we provide an in-depth analysis of configuration parameters that allow for a speed-up of the answering time and a decrease of storage space needed for query processing. Thereupon, we provide configuration recommendations for process mining with SQL on relational databases.

Process mining, which aims to mine a high-quality process model from event log, provides a powerful tool to support the design, enactment, management, and analysis of operational business processes. However, the task is not easy because the algorithm needs to discover various complex process structures, handle noisy and incomplete event logs and balance multiple performance indicators. In this paper, a novel algorithm (called PSOMiner) for process mining is proposed, which consists of a discrete particle swarm optimization algorithm and guided local mutation. The former is in charge of searching the solution space of causal matrix and the latter is used to help the algorithm skip out the local optimum when it suffers from premature. A fine-grained scoring strategy which used to assign a score to each position of a particle (i.e. causal matrix) is presented to direct the mutation. The experiments were performed on 28 synthetic event logs with/without noise and 4 real-life event logs, and three classical algorithms of process mining (ETM, Hybrid ILP Miner, HM) were chosen for comparison. The results show that (1) PSOMiner achieved the best f-score on 25 synthetic event logs; (2) The average f-score of PSOMiner is 0.825 on 4 real-life event logs, which is superior to ETM whose average f-score is 0.703.

It is difficult for tunnel boring machine (TBM) operators to respond for safe and high-efficient construction without accurate reference parameters such as the TBM thrust. A new hybrid model (MRFO-AT-TELM) combining an improved two-hidden-layer extreme learning machine (AT-TELM) and manta ray foraging optimization (MRFO) algorithm is proposed to predict TBM thrust with 12 selected input featuring parameters. The affine transformation (AT) activation function is used to improve the performance of TELM. Input weights and bias of AT-TELM are optimized using the MRFO algorithm. The performance of the proposed model is validated with TBM construction data collected from the Yin-Song Project in China and compared with other models. Input data of the first 30, 60, and 90 seconds of the rising period are analyzed. Results show that the proposed model is superior to the other models and with 90-second data as input outperforms that with 30 and 60-seconds data. The proposed model and the selected input features are validated in a new project. The thrust prediction model can be embedded into the TBM construction intelligence system and thus help improve construction efficiency.

Time prediction based on process mining

This paper is dedicated to a tool whose aim is to facilitate process mining experiments and evaluation of the repair algorithms. Process mining is a set of approaches which provides solutions and algorithms for discovery and analysis of business process models based on event logs. Process mining has three main areas of interest: model discovery, conformance checking and enhancement. The paper focuses on the latter. The goal of enhancement process is to refine an existing process model in order to make it adhere event logs. The particular approach of enhancement considered in the paper is called decomposed model repair. Although the paper is devoted to the implementation part of the approach, theoretical preliminaries essential for domain understanding are provided. Moreover, a typical use case of the tool is shown as well as guides to extending the tool and enriching it with extra algorithms and functionality. Finally, other solutions which can be used for implementation of repair schemes are considered, pros and cons of using them are mentioned.

The aim of process discovery is to build a process model from an event log without prior information about the process. The discovery of declarative process models is useful when a process works in an unpredictable and unstable environment since several allowed paths can be represented as a compact set of rules. One of the tools available in the literature for discovering declarative models from logs is the Declare Miner, a plug-in of the process mining tool ProM. Using this plug-in, the discovered models are represented using Declare, a declarative process modelling language based on LTL for finite traces. In this paper, we use a combination of an Apriori algorithm and a group of algorithms for Sequence Analysis to improve the performances of the Declare Miner. Using synthetic and real life event logs, we show that the new implemented core of the plug-in allows for a significant performance improvement.

Energy storage has the characteristics of strong flexibility and fast response, which can effectively alleviate the load fluctuation, voltage instability, harmonic pollution and other problems caused by the high proportion of new energy access. In this paper, a power system energy storage locating and sizing based on bi-level model and improved manta ray foraging optimization algorithm is proposed. First of all, a bi-level locating and sizing model is established to minimize energy storage investment cost, daily average voltage fluctuation and daily average load fluctuation. Secondly, the elite reverse learning strategy and adaptive tumbling factor are used to improve the manta ray foraging optimization algorithm, and the improved algorithm is used to solve the model. Finally, the proposed method is simulated by taking the IEEE39 node system with high proportion of new energy as an example. The results prove that the proposed method has faster operation speed and higher accuracy, and the locating and sizing selection scheme can effectively reduce the system investment cost and significantly reduce the system voltage and load fluctuation.

Quantum-Inspired Manta Ray Foraging Optimization Algorithm for Automatic Clustering of Color Images