research article

Understanding Machine Learning Algorithms into Multiproduct Batch Plant Design of Protein Production

Youness El Hamzaoui*

Faculty of Engineering, Autonomous University of Carmen, Mexico

*Corresponding author: Youness El Hamzaoui, Faculty of Engineering, Autonomous University of Carmen, Mexico

Received Date: 15 January, 2020; Accepted Date: 24 January, 2020; Published Date: 31 January, 2020

Citation: Hamzaoui YE (2020) Understanding Machine Learning Algorithms into Multiproduct Batch Plant Design of Protein Production. J Pharma Pharma Sci: 4: 184. DOI: 10.29011/2574-7711.100084

The problem of optimization of Multiproduct Batch Plant Design (MBPD) in chemical engineering systems where the design variables are the size of the equipment elements and the setting of operating conditions. The application is a multiproduct batch plant for the manufacture of four recombinant proteins as insulin, chymosin, vaccine and protease. However, addressing an important class of optimization problems handed over the serious combinatorial aspect of the complication. The procedure implemented consists in using machine learning algorithms, in order to minimize the investment cost and find out the number and size of parallel equipment units in each stage. The calculation results (investment cost, number and size of equipment, computational time, CPU time, idle times in plant, production rate, annual demand rate, setup cost, holding cost, variable cost, selling price, inspection rate, return cost, penalty cost, screening cost) obtained by metaheuristics machine learning tools are better than mixed integer no linear programming. This approach can facilitate the manufacturers of pharmaceutical drug to get an optimal design and makes up a remarkably suggested plan for having a benefit of efficient results.

Keywords: Batch Plant Design; Chemical Engineering Optimization; Machine Learning Algorithms; Mathematical Modeling

1. Introduction

Pharmaceutical researchers and biotechnology companies are devoted to developing medicines, such as: therapeutic proteins, human insulin, vaccines for hepatitis, food grade protein, chymosin detergent enzyme, and cryophilic protease. This allows patients to live longer, heathier, and more productive. Within this context, there is a high degree of consensus in the biomanufacturing industry that product quality, customer service, and cost efficiency are fundamental for success. The pharmaceutical industry must join its effort with government and the health professions to seek new, innovative, and cost effective approaches in the development process. However, the pharmaceutical process is characterized by a membership function involved in the field of chemical engineering.

Nevertheless, for understanding a chemical engineering system, we have to go back to the mathematical modeling, as we know the mathematical modeling is a powerful tool to solve different problems which arise in chemical engineering optimization. Problems as designing a plant, determining the number of units for a specific task, assigning raw materials to different production processes and deciding the production planning or production targets are some of the issues that can be solved through mathematical modeling. In other words, the mathematical formulations are used to make decisions at different levels, from the synthesis and design of the process up to its operation and scheduling [1]. In spite of that, precisely in recent years, there has been an increased interest development of systematic method for the design of batch process in chemicals, food products, and pharmaceutical industries. Basically, batch plants are composed of items operating in a discontinuous way. Each batch then visits a fixed number of equipment items, as required by a given synthesis sequence so called production recipe. That means, the design of batch plants requires involving how equipment may be utilized. In addition, the optimal design of a multiproduct batch chemical process involves the production requirement of each product and the total production time available for all products has been considered. The number and size of parallel equipment units in each stage as well as the location and size of intermediate storage are to be determined in order to minimize the investment cost.

Many works in the literature on batch process design are based on expressions that relate the batch sizes linearly with the equipment sizes [2], made a comprehensive framework for optimal design of batch plants. Dietz, Azzaro-Pantel, Pibouleau & Domenech [3], developed an approach of multiobjective optimization for multiproduct batch plant design under economic and environmental considerations. Ponsich, Azzaro-Pantel, Domenech & Pibouleau, illustrated some guidelines for genetic algorithms implementation in MINLP batch plant design problem [4]. Dietz, Aguilar-Lasserre, Azzaro-Pantel, Pibouleau & Domenech, presented of fuzzy multiobjective algorithm for multiproduct batch plant [5]. In addition, they’ve used genetic algorithms to solve multiobjective optimization problem with an application to optimal batch plant design in process system engineering. Aguilar-Lasserre, Bautista, Ponsich & González Huerta, developed an AHP-based decision making tool for the solution of multiproduct batch plant design problem under imprecise demand [6]. Aguilar-Lasserre, Giner, Azzaro-Pantel, Guillermo, Constantino, Pibouleau & Rubén, illustrated the problem of the optimal design of batch plants with imprecise demands using concepts of fuzzy logic [7]. Borisenko, Kegel & Gorlatch, developed and performed a parallel algorithm for finding optimal design for multiproduct batch plants [8].

In the conventional optimal design of a multiproduct batch chemical plant [9], a designer specifies the production requirements for each product and total production time for all products [10]. The number required of volume and size of parallel equipment units in each stage is to be determined in order to minimize the investment cost.

The case of study is a multiproduct batch plant for the production of proteins taken from the literature, we will only consider multiproduct batch plants, which means that all the products follow the same operating steps [11,12], the structure of the variables are the equipment sizes and number of each unit operation that generally take discrete values. Generally, optimization of multiple parameters is an arduous and time consuming task. In this context, we emphasize referring to the work of Montagna, et al. [13], and Asenjo, et al. [14] about the strategy based on monoproduct campaigns was assumed, even when considering the design of multiproduct batch plant. Therefore, machine learning applications are everywhere, from self-driving cars, spam detection, document search, trading strategies, and even speech recognition. This makes machine learning suitable for the era of big data era and data science, especially in pharmaceutical and pharmacological sciences. The main challenge is how to convert data to see what is possible.

The aim of this work is to solve the multiproduct batch plant design problem using (PSA) and (GAs), respectively. The model presented is general, it takes into account all the available options to increase the efficiency of the batch plant design: unit duplication in-phase and out-phase and intermediate storage tanks.

We have found out that PSA performs effectively and gives a solution, but we would like to solve the problem more effectively, that’s why we proposed to apply GAs, an intelligent problem-solving method that has demonstrated its effectiveness in solving combinatorial optimization problem, and satisfactory results are obtained [15].

The paper is organized as follows, section 2 is devoted to the materials and methods including the system description and experimental data, problem statement, model equations and the methodology. While, the results and discussions are handling and reported in section 3. Finally, the conclusions of this work are drawn.

2. Materials and Methods

2.1. System Description and Experimental Data

The case study, taken from the literature, is a multiproduct batch plant for the production of proteins [16]. This example is used as a test bench since it provides models describing the unit operations involved in the process. The batch plant involves eight stages for producing four recombinant proteins, on one hand, two therapeutic proteins, human insulin (A) and vaccine for hepatitis (B) and, on the other hand, a food grade protein, chymosin (C), and a detergent enzyme, cryophilic protease (D). Figure 1 is the flowsheet of the multiproduct batch plant considered in this study.

All the proteins are produced as cells grow in the fermenter. It is hardly necessary to say that the number of intermediate storage tanks is an important constituent of our process: Three tanks have been selected: the first after the fermenter, the second after the first ultrafilter, and the third after the second ultrafilter.

Vaccines and protease are considered to be intracellular. The first microfilter is used to concentrate the cell suspension, which is then sent to the homogenizer for the second microfilter, which is used to remove the cell debris from the solution proteins. The first ultrafiltration step is designed to concentrate the solution in order to minimize the extractor volume. In the liquid–liquid extractor, salt concentration (NaCl) is used as solution in order to minimize the extractor volume. In the liquid–liquid extractor, salt concentration (NaCl) is used to first drive the product to a Poly-Ethylene-Glycol (PEG) phase and again into an aqueous saline solution in the back extraction. The second ultrafiltration is used again to concentrate the solution. The last stage is chromatography, during which selective binding is used to better separate the product of interest from the other proteins.

Insulin and chymosin are extracellular products. Proteins are separated from the cells in the first microfilter, where cells and some of the supernatant liquid stay behind. To reduce the amount of valuable products lost in the retentate, extra water is added to the cell suspension. The homogenizer and the second microfilter for cell debris removal are not used when the product is extracellular. Nevertheless, the first ultrafilter is necessary to concentrate the dilute solution prior to extraction. The final step of extraction, second ultrafiltration, and chromatography are common to both the extracellular and intracellular products. In Table 1 we make an estimation of production targets and product prices [17-19].

2.2. Problem Statement

The model formulation for DMBP’s problem approach adopted in this section is based on Montagna, et al. [16]. It considers not only treatment in batch steps, which usually appear in all types of formulation, but also represents semi continuous units that are part of the whole process (pumps, heat exchangers, etc). A semi-continuous unit is defined as a continuous unit alternating idle times and normal activity periods. Besides, this formulation takes into account mid-term intermediate storage tanks, the obligatory mass balance at the intermediate storage stage, which is one of the most efficient strategies to decouple bottlenecks in batch plant design. They are just used to divide the whole process into subprocesses in order to store an amount of materials corresponding to the difference of each sub-process productivity. In this section we describe the unit models from a conceptual standpoint and also the procedure to derive the data needed for solving the mathematical model. These data are summarized in Tables 2 and 3.


 

Similarly, vaccine, chymosine, and cryophilic protease were estimated to be 0.1, 0.15, and 0.2 of total proteins of the biomass, respectively. The batch stage description is completed by estimating a processing time Tij for stage when producing product i. For the fermenter, we estimate Tij for all products, which includes time for charging, cell growth, and discharging.

This model of batch stages given by constraint (1a) is the simplest one. Its level of detail suffices for the fermenter and the extractor. These units are truly batch items chat hold the load to be processed and whose operations are governed by kinetics, and hence, the operating time does not depend on the batch size.

As a first approximation for the extractor, we take a phase ratio of (1b) for all products. Therefore, the required extractor volume is twice the inlet batch volume, while the inlet and outlet aqueous saline batches are of the same volume. It is also assumed, as a result of preliminary balances, that this operation reduces the total amount of proteins to about twice the amount of the target protein. With respect to the kinetic effects we take as first estimates [23] the following times: 15 min stirring to approach phase equilibrium, 30 min settling to get almost complete disengaging of the phases, and 20 min for charging and discharging. A special consideration must be done in the case of the microfiltration, homogenization, and ultrafiltration stages. Although the mathematical model considers them batch stages, their corresponding equipment consists of holding vessels and semicontinous units that operate on the material that is recirculated into the holding vessel. The batch items are sized as described before. For example, for the homogenizer processing cryophilic protease, we estimated that the fermentor broth is concentrated 4 times up to 200kg/mat microfilter 1 and considered a yield of 1 because the intracellular protease is fully retained at the microfilter. Then the size factor of the homogenizer vessel is 4 times smaller than the fermenters, i.e., Sij=0.08 mprotease. The sizing equation for semicontinuous items can also be found in the general batch processes literature [24]:




The general batch processes literature considers semicontinuous units to work in series with batch units so that their operating time are the times for filling or emptying the batch units. However, in the process considered, pumps are the only semicontinuous units, which transfer batches between the units. As the pumps cost does not have a relevant impact on the plant design, they were not explicitly modeled. The times for filling and emptying batch items were estimated and included in the batch cycle times. On the other hand, the process does have special semicontinuous units with an important economic impact on the cost. They are the homogenizer and ultrafilters, but their operating time is the batch processing time of the respective stage. Their mathematical model has been introduced by Salomone and Iribarren, 1994. A size factor for the batch item and a time expression for the stage that depends on both the batch size and the size of the semicontinuous item are as follows:



This ratio is estimated from a mass balance taking into account that the ultrafilters are used for a water removal from solutions up to 50g/L of total proteins. Ultrafilters are used to reduce the volume required at the liquid extractor and the chromatographic column. The upper bound on concentration is a constraint that avoids protein precipitation. The microfilter model is quite similar to that of the ultrafilter, but there are two batch items associated to them instead of one, the retentate and the permeate vessels, plus the semicontinuous item area of filtration. For microfilter 1 a fixed permeate flux of 200L/m2h is adopted. For extracellular insulin and chymosin, we estimate a total permeate (feedwater plus make up water) twice the feed, while for intracellular protease and vaccine we estimate it in 75% of the feed (the retentate is concentrated four times). For microfilter 2 a fixed permeate flux model is also used. In this case, the flux is smaller than the one in microfilter 1 because the pore size to retain cell debris is smaller than the one for whole cells. As a first estimation we take 100L/m2h and a total permeate (feed plus make up water) twice the feed. With respect to the chromatographic column, an adsorptive type chromatography is considered, with a binding capacity of 20kg/m3 column packing. The size factor of this unit is the inverse of that binding capacity. As a first approximation, a fixed total operating time of 0.5h was estimated for loading, eluting, and washing regeneration.

Finally, the stage model is completed with a cost model that expresses the cost of each unit as a function of its size, in the form of a power law. These expressions are summarized in Table 4, with most of the cost data taken from Petrides, et al. [19].

2.3. Model Equations

The mathematical optimization model for designing the multiproduct batch plant is described in this section. The model includes the stage models described in the previous section plus additional constraints that are explained in this section. The plant consists of M batch stages (in our case 8 batch stages). Each stage  has a size V1(m3), and more than one unit can be installed in parallel. They can work either in-phase (starting operation simultaneously) or out of phase (starting times are distributed equally spaced between them). The duplication in phase is adopted in case the required stage size exceeds the specific upper bound. In this case Gunits are selected, splitting the incoming batch into Gj smaller batches, which are processed simultaneously by the Gunits. After processing, the batches are added again into a unique outgoing batch. Otherwise, duplication out-of-phase is used for time-limiting stages, if a stage has the largest processing time, then it is a bottleneck for the production rate. Assigning Mj units at this stage, working in out of phase mode, reduces the limiting processing time and thus increases the production rate of the train. For this case, the batches coming from the upstream stages are not split. Instead, successive batches produced by the upstream stage are received by different units of stage j, which in turn pass them at equally spaced times onto the downstream batch stage. The allocation and sizing of intermediate storage has been included in the model to get a more efficient plant design. The goal is to increase unit utilization. The insertion of a storage tank decouples the process into two subprocesses: one upstream from the tank, and the other downstream. This allows the adoption of independent batch sizes and limiting cycle times for each subprocess.

Therefore, the previously unique Bis changed to batch sizes Bij defined for product in stage j. Appropriate constraints adjust the batch sizes among different units. The objective is to minimize the capital cost of the plant. The decision variables in the model are as follows: At each batch stage the number of parallel units in phase and out of phase and their size, and the installation or absence of intermediate storage between the batch stages and their size. The plant is designed to satisfy a demand of Qi(kg) each product i, for the product considered, within a time horizon H(h).

In summary, the objective function to be optimized is

 











2.4. Methodology

Between 1960s and 1970s witnessed a tremendous development in the size and complexity of industrial organizations. Administrative decision-making has become very complex and involves large numbers of workers, materials and equipment. A decision is a recommendation for the best design or operation in a given system or process engineering, so as to minimize the costs or maximize the gains [28]. Using the term "best" implies that there is a choice or set of alternative strategies of action to make decisions. The term optimal is usually used to denote the maximum or minimum of the objective function and the overall process of maximizing or minimizing is called optimization. The optimization problems are not only in the design of industrial systems and services, but also apply in the manufacturing and operation of these systems once they are designed. Including various methods of optimization, we can mention: MINLP, Particle Swarm Optimization and Genetics Algorithms.

2.4.1. Particle Swarm Algorithms

The PSA is a population-based optimization algorithm, which was inspired by the social behavior of animals such as fish schooling and birds flocking, it can solve a variety of hard optimization problems. It can handle constrains with mixed variables requiring only a few parameters to be tuned, making it attractive from an implementation viewpoint [29]. In PSA, its population is called a swarm and each individual is called a particle. Each particle flies through the problem space to search for optima. Each particle represents a potential solution of solution space; all particles form a swarm. The best position passed through by a flying particle is the optimal solution of this particle and is called pbest, and the best position passed through by a swarm is considered as optimal solution of the global and is called gbest. Each particle updates itself by pbest and gbest. A new generation is produced by this updating. The quality of a particle is evaluated by value the adaptability of an optimal function. In PSA, each particle can be regard as a point of solution space. Assume the number of particles in a group is M, and the dimension of variable of a particle is N. The ith particle at iteration k has the following two attributes:


 


2.4.2. Genetic Algorithms Approach

GA, proposed in this paper based on the work of Wang, et al. [30], are related to the mechanics of natural selection and natural genetics. They combine the survival of the fittest among string structures with a structured yet randomized information exchange to form search algorithms with some of the innovative flair of human search. In every generation, a new set of individuals (strings) is created using bits and pieces of the fittest of the old individuals; while randomized, a GA are no simple random walk. They efficiently exploit historical information to speculate on new search points with expected improved performance [30]. According to Wang, et al. [30], the canonical steps of the GA can be described as follows:





3. Results and Discussions

The problem could be formulated as the minimization of the investment cost for equipment and storage tanks. Given that the problem modeled has non linear objective function. For the purpose of optimization problem, the model developed has been solved with PSA and Gas Matlab Toolbox respectively, which is included in the Matlab optimization modeling software, using the data shown in Tables 1, 2, 3, 4. A horizon time of 6000 h has been considered.

Table 7 shows the best, the average and the worst among the final fitness values and the related standard deviation obtained in the 30 runs of PSA and GA, respectively. 

It is clear from the summary of the results shown in Table 7, that the performance of both PSA and GA produce adequate values regarding the cost for equipment and storage tanks. However, GA performs better than the PSA in terms of the average and the worst fitness values and the standard deviation. Table 7, also, shows the best final solution found in the 30 runs of PSA and GA. According to our knowledge, the case study about the optimal design of protein production plant has been taken from Montagna, et al. [16]. However, they solved the problem using rigorous mathematical programing (MINLP), their model includes 104 binary variables and has been convexified using the transformation proposed by Kocis and Grossman. The MINLP model has been solved using DICOPT++, which is included in the GAMS optimization modeling software. The algorithm implemented in DICOPT++ relies on the Outer Approximation/Equality Relaxation/Augmented Penalty (OA/ER/AP) method. The OA/ER/AP solution method consists of the decomposition of the original MINLP problems into a sequence of two subproblems: a Non Linear Programming (NLP) subproblem and a Mixed Integer Linear Programming (MILP) subproblem also known as the Master problem, which is solved to global optimality (minimize the caplital cost $829,500). However, in previous work of Montagna, et al [16], their model needed a long computational time (more than 86400 seconds) and require several initial values to the optimization variables, they also showed in their paper that the behavior of the demand was completely deterministic.

Whilst, this assumption does not seem to be always a reliable representation of the reality, since in practice the demand of pharmaceutical products resulting from the batch industry is usually variable. Simulations outcomes were then compared with experimental data in order to check the accuracy of the method. The error from the optimal solution is given by:


In this research, Xexf is considered to be the optimal solution founded by Montagna (Plant cost $829,500), where the equation 19 is a criterion to confirm the optimal values. Table 8 presents the results obtained in different optimization runs for multiproduct batch plant design. For each simulator run, the average numerical effort spent on solving the problem on LINUX System, Intel ® D, CPU2.80 Ghz, 2.99 of RAM. Table 8 shows plant cost, % from optimal solution and CPU time obtaining during 30 runs. PSA and GA performed effectively and give a solution within 10 and 0.5% of the global optimal $912,450 and $833,647, respectively. Furthermore, the important feedback could be taken from Table 8, is the GA results in a faster convergence than PSA and the MINLP algorithm. In addition, the GA is so close to the global optimal of MBPD (0.5% from optimal solution) and provides also an interesting solution, in terms of quality as well as of computational time as illustrated in Table 8, while Table 9 presents the sizes for the units involving a set of discrete equipment structure given by PSA. The inconvenience of this configuration is just stopped at 6000h with risk of failing to fulfill the potential future demand coming from a fluctuation of the market.

In order to show how the evolution process is going on for both PSA and GAs, respectively, the convergence of the best fitness values is shown in Figure 1. The convergence rate of objective function values as a function of generations for both PSA and GAs is shown in Figure 1, where for clarity only 1000 generations are shown. It is clear from this figure that, for the optimization problem considered, GAs decrease rapidly and converge at a faster rate (around 500 generations) compared to that for PSA (about 800 generations), from which it is clear that GAs seem to perform better compared to PSA. So, for the present problem the performance of the GAs is better than PSA from an evolutionary point of view.

To compare the computational time, the swarm/population size is fixed to 200 for both PSA and GAs algorithms. Whereas, the generation number is varied. Simulation were carried out and conducted on LINUX System, Intel (R) D, CPU 2.80 Ghz, 2.99 of RAM Computer, in the MATLAB 7.0.1 environment. Here the result in the form of graph is shown in Figure 1. It is clear from Figure 1 that the computational time for GAs is very low compared to the PSA optimization algorithm. Further, it can also be observed from Figure 11 that in case of GAs the computational time increases linearly with the number of generations, whereas for PSA the computational time increases almost exponentially with the number of generations. The higher computational time for PSA is due to the communication between the particles after each generation. Hence as the number of generations increases, the computational time increases almost exponentially.

Table 9 presents the sizes for the units involving a set of discrete equipment structure given by PSA. The inconvenience of this configuration is just stopped at 6000 hours with risk of failing to fulfill the potential future demand coming from a fluctuation changing of the market.

On the other hand, the calculation of the structure of equipment using GA is illustrated in Table 10. The total production time, also, computed by GA is 5491.12 hours to fulfill the eventual increase of future demand caused by market fluctuation. In addition, the GA results in a faster convergence. However, the equipment structure showed by PSA is very expensive. Furthermore, the PSA approach has the disadvantage of long CPU time.

At the same time as, the GA allow the reduction of the idle time to the stage, in any way, Table 11 and Table 12 show the idle times obtained by PSA and GA respectively.

However, some observations about some important aspects in our implication of GAs and some problems in practice: the most important of all is the method of coding, because the codification is very important issue when a genetic algorithm is designed to dealing with combinatorial problem, also of the characteristics and inner structure of the DMBP.

The commonly adopter concatenated, multi-paramer, mapped, fixed point coding are not effective in searching for the global optimum. According to the inner structure of the design problem of multiproduct batch that gives us some clues for designing the above mixed continuous discrete coding method with a four-point crossover operator. As is evident from the results of application, this coding method is well fit for the proposed problem. Another aspect that affects the effectiveness of our Genetic Algorithms procedure considerably is crossover.

Corresponding to the proposed coding method, we adopted a four-point crossover. It is commonly believed that multipoint crossover is more effective than the traditional one-point crossover method. It is also important to note that the selection of crossover points as well as the way to carry out the crossover should take in account the bit string structure, as is the case in our codification.

One problem in practice is the premature loss of diversity in the population, which results in premature convergence, because premature convergence is so often the case in the implementation of GA according to our calculation experience. Our experience makes it clear that the Elitism parameter can solve the premature problem effectively and conveniently. However, a numerical calculation of the model under machine learning approach is examined in table 13.

In order to further explain the effects of these algorithms on solving the MBPD problem, the variance analysis was performed. Each of the PSA and GA algorithms was run 30 times. The Minitab software was used to analyze the results. Therefore, the results are given in tables 14 and 15.

Table 14 indicates that, the mean square deviation between groups (SDB) is 779.895. The mean square deviation within groups (SDI) is 50.392. The test statistic F = 15.477. If significance level α = 0.05, then the critical value

2.92≤ Fα(3.36)≤2.84. Thus, F> Fα(3.36) indicating that the difference between the average figures is significant, that is, the performance difference of algorithms is significant.

Nevertheless, these techniques are not a panacea, despite their apparent robustness, there are control “parameters” involved in these metaheuristics and appropriate setting of these parameters is a key point for success.

6. Conclusions

Techniques such as PSA and GA are inspired by nature, and have proved themselves to be effective solutions to optimization problems. We applied Genetic Algorithms with an effective mixed continues discrete coding method with a four crossover point to solve the problem of DMBP. GA perform effectively and give a solution within 0.5% of the global optimum. Whilst, it is observed that, in terms of computational time, the GAs approach is faster. The computational time increases linearly with the number of generations for GA, whereas for PSA the computational time increases almost exponentially with the number of generations, interpreting that, the higher computational time for PSA is due to the communication between the particles after each generation. Furthermore, the results provided by GA are much better with respect to PSA. In this paper, the GA gave us the highest efficiency and justifies its use for solving nonlinear mathematical models. Therefore, this work provides an interesting decision/making approach to improve the design of multiproduct batch plants under conflicting goals.

7. Acknowledgements

The author express their gratitude, appreciation and acknowledgements to the financial support provided by PRODEP (Programa para el Desarrollo Profesional Docente para el tipo Superior) under the Research Project 511-6/18-8724.


Figure 1: Multiproduct batch plant for protein production.


Table 1: Product prices and demands.



Table 2: Size factors Sij (r, retentate; p, permeate).



Table 3: Time factors Tij [Bi (kg)].



Table 4: Cost of equipment (U.S.dollars).



Table 5: The parameters used for running GA and PSA.



Table 6: Intermediate storage cost coefficients and size factors.



Table 7: Comparison of results for 30 runs between PSA and GA.