International Journal of Trend in Scientific

Research and Development (IJTSRD)

International Open Access Journal
ISSN No: 2456 - 6470 | www.ijtsrd.com | Volume - 2 | Issue – 1

Agent-Based
Based Modeling: Methods and Techniques for
Scheduling Industrial Machine Operation

Chiagunye Tochukwu Inyiama Hyacinth Aguodoh Patrick

Department of Computer Department of Electronic and Department
epartment of Computer
Engineering, Michael Okpara Computer Engineering, Nnamdi Engineerin,
Engineerin Michael Okpara
University of Agriculture, Azikiwe University, Awka University of Agriculture,
Umudike Umudike

ABSTRACT
Owing to its impact on the industrial economy, job parts end up in a wide variety
ty of products sold in the
shop scheduler and controller are vital algorithms for US and elsewhere. These factories employ roughly
r
modern manufacturing processes. Hence a production over 3 million people and ship close to $ 7 billion
scheduling and control that performs reactive (not worth of products every year. The vast majority of
deterministic) scheduling and can make decision on these factories are called “job shops”, meaning that
which job to process next based solely on its partial the flow of raw and unfinished goods through them is
(not central) view of the plantnt becomes necessary. completely random. Over the years, the behavior and
This requirement puts the problem in the class of performance of these job shops have been the focus of
agent based model (ABM). Hence this work adopt adopts considerable attention in Operations Research (OR)
an alternative view on job-shop
shop scheduling probl
problem literature.
where each resource is equipped with adaptive agent
that, independent of other agents makes job Manufacturing industries are facing a growing and
dispatching decision based on its local view of the rapid change. Major trends like globalization,
plant. A combination of Markov Chain instruments customer orientation and increasing
reasing market dynamics
and agent oriented analysis is used in the analysis of lead to a shift in both managerial and manufacturing
the
he proposed agent based model (ABM) for the job principles: enterprises have to become more flexible,
shop scheduling problem. The Markov Chain open, fast, effective, self-organized,
organized, decentralized, to
approach allows a rigorous analysis of the ABM. It sum it up: agile [2].
]. Manufacturing serves as a basic
provides a general framework of aggregation in agent
agent- function for any agile enterprise. The call for agility
based and related computational models by making challenges the shop floor with several problems.
use of Markov ov Chain aggregation and lumpability With an increasing occurrence of changes and
theory in order to link the micro and the macro level dominating customer demand, management of
of observation. Simulated annealing technique is used manufacturing processes and the coordination of the
for carrying out the optimization modeling for the multifarious resources, i.e., machines, materials,
ABM. information, knowledge and humans, becomes a core
task of shop floor scheduling and control algorithm
Keywords: Agent-Based, Scheduling
Scheduling, Industry, [2].
Machine Operation, Markov Chain
An important issue in a manufacturing environment is
1.1 Introduction the improvement of resource utilization. A classical
way of achieving
ing improved resource utilization is by
In the United States alone, there are over 40 40,000 using scheduling algorithms [3
3]. As defined by Baker
factories producing metal-fabricated
fabricated parts [[1]. These [4] scheduling is concerned with the problem of

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec

Dec 2017 Page: 465
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
assigning a set of jobs to resources over a period of publications argued that agent-based modeling is used
time. Effective scheduling plays a very important role because only agent-based model can explicitly
in today’s competitive manufacturing environment. incorporate the complexity arising from individual
Performance Criteria such as machine utilization, behavior and interactions that exist in the real-world.
manufacturing lead times, inventory costs, meeting [9] contributed to interweaving Markov Chains and
due dates, customer satisfaction, and quality of agent-based modeling. [10] worked on and
products are all dependent on how efficiently the jobs represented ten well-known simulation models as a
are scheduled in the system [5]. Hence, it becomes time homogenous Markov Chain. The author’s main
increasingly important to develop effective scheduling idea is the formulation of the system configuration as
approaches that help in achieving the desired the state space of the Markov Chain.
objectives.
[11] formulated the JSP using integer programming.
The scheduling and planning of production order have This technique involved assuming that each job
an important role in the manufacturing system. The consist of m operations and must pass through each
diversity of products, increased number of orders, the machine exactly once. All machines are available at
increased number and size of workshops and time zero. Furthermore, the total number of sub lots
expansion of factories have made the issue of is given and consistent sub lot sizes are considered.
scheduling production orders more complicated, Concerning the limitation of this technique, [12]
hence the traditional methods of optimization are pointed out that the problem formulation used does
unable to solve them [6][7]. not recognize the physical environment of the shop
floor domains where interference not only leads to
As can be deduced from these techniques, most readjustment of schedules but also imposes physical
approaches to job-shop scheduling assume complete conditions to minimize them. Still, difficulties in the
task knowledge and search for a centralized solution. formulation of material flow constraints as
These techniques typically do not scale with problems mathematical inequalities and the development of
size, suffering from an exponential increase in generalized software solutions have limited the use of
computation time. The centralized view of the plant these approaches.
coupled with the deterministic algorithms
characteristic of these schedulers do not allow the Furthermore, although efficient bonding and pruning
manufacturing processes to adjust the schedule (using procedures have been developed to speed up the
local knowledge) to accommodate disturbances such search, this is still a very computational intensive
as machine breakdowns. Hence a production procedure for solving large scheduling problem. [13]
scheduling and control that performs reactive (not proposed a methodology for solving the job shop
deterministic) scheduling and can make decision on problem based on the decomposition of mathematical
which job to process next based solely on its partial programming problems that used both Benders-type
(not central) view of the plant becomes necessary. [14] and Dantzig/wolfe-type [15] decompositions.
This requirement puts the problem in the class of The methodology was part of closed-loop, real-time,
agent based model (ABM). Hence this work adopts two-level hierarchical shop floor control system. The
an alternative view on job-shop scheduling problem top-level scheduler (i.e., the supremal) specified the
where each resource is equipped with adaptive agent earliest start time and the latest finish time for each
that, independent of other agents makes job job. The lower level scheduling modules (i.e., the
dispatching decision based on its local view of the infimals) would define these limit times for each job
plant. by detailed sequencing of all operations. A multi-
criteria objective function was specified that include
2.0 Literature Review tardiness, throughput, and process utilization cost.
The limitations of this methodology stem from the
Scheduling, understood to be an important tool for inherent stochastic nature of job shops and the
manufacturing and engineering, has a major impact on presence of multiple, but often conflicting, objectives
productivity of a process [8]. In manufacturing, the made it difficult to express the coupling constraints
purpose of scheduling is to minimize the production using exact mathematical relationships. This made the
time and cost, by telling a production facility what to schedule not to converge. Furthermore the rigid
make with which staff, and on which machine. Cited

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec 2017 Page: 466
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
centralization of the scheduler made it not able to scheduling problem. The Markov Chain approach
adjust to disturbances at the shop floor. allows a rigorous analysis of the ABM. It provides a
general framework of aggregation in agent-based and
[16] evaluated the use of MRP or MRP-11 to create a related computational models by making use of
medium-range scheduler. MRP system’s major Markov Chain aggregation and lumpability theory in
disadvantages are not only rigidity and the lack of order to link the micro and the macro level of
feedback from the shop floor, but also the tremendous observation.
amount of data that have to be entered in the bill of
materials and the fact that the model of the The starting point is a Microscopic Markov Chain
manufacturing system and its capacity are excessively description of the dynamical process in complete
simple. As can be deduced from these techniques, correspondence with the dynamic behavior of the
most approaches to tackle job-shop scheduling agent model which is obtained by considering the set
problem assume complete task knowledge and search of all possible agent configurations as the state space
for a centralized solution. These techniques typically of a huge Markov Chain. This is referred to as micro
do not scale with problem size, suffering from an chain. The explicit micro formulation enables the
exponential increase in computation time. The application of the theory of Markov Chain
centralized view of the plant coupled with the aggregation-namely, lumpability-in order to reduce
deterministic algorithms characteristic of these the state space of the micro chain and relate
schedulers do not allow the manufacturing processes microscopic deceptions to a macroscopic formulation
to adjust the schedule (using local knowledge) to of interest.
accommodate disturbances such as machine break Simulated annealing technique is used for carrying
downs. Hence a production scheduling and control out the optimization modeling for the ABM. Some of
that performs reactive (not deterministic) scheduling the conditions for asymptotic convergence, as for
and can make decision on which job to process next instance, the infinite length of Markov Chains, cannot
based solely on its partial (not central) view of the be met in practice [1]. Hence, in any finite
plant becomes necessary. This requirement puts the implementation, choice have to be made with respect
problem in the class of agent based model (ABM). to the parameter:
Hence this work adopts an alternative view on job - The length of the Markov Chain
shop scheduling problem where each resource is - The initial value of the control parameter
equipped with adaptive agent that independent of - The decrement rule of the control parameter
other agents, makes job dispatching decision based on - The final value of the control parameter
its local view of the plant.
Such a choice is usually referred to as cooling
3.0 Research methodology: scheduler or an annealing schedule [2], hence the use
of simulated annealing for the global optimization of
A combination of Markov Chain instruments and
the ABM for the proposed JSP.
agent oriented analysis is used in the analysis of the
proposed agent based model (ABM) for the job shop

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec 2017 Page: 467
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
3.2 Proposed System

Update orders; resources, agent address

Update orders, resources

Register schedule, update schedule

Scheduler Shop Floor Neighbor

Agent Execute schedule Control Agent Agent
Data

Register Request
Register new Resources, Order Manufacturing
update machine status
Operation
Machine Control
(workstation) Order
Agent Agent

Request production capacity

Fig. 1: Block Diagram of the Proposed ABM Framework for Solving the JSP

Fig. 1 presents the shop for scheduler agent, shop agent that makes decision on which job to process
floor control (SFC) agent, order agent and the next based solely on its partial view of the plant. For
machine (or work station) agent. This figure shows executing a schedule, the machine (workstation) agent
the interaction between the scheduler agent and other needs to register itself with the shop floor scheduler
agents. agent. Further on, the machine agent shall inform the
scheduler agent and SFC agent when it is available,
First, the agents mentioned above needs a model of occupied, or unavailable in the (near) future. When it
the surrounding agents. Agents have to register becomes available again, it notifies this event to the
themselves with the SFC agent. Second, resources scheduler agent. The scheduler agent can explicitly
like a machine (or workstation) need to notify the ask for the availability of the machine. The SFC
shop control agent when they have spear productivity agent, considering all available information, will
capability. The order agent sends request for allocate a machine (workstation) to a task. It will
production capacity (to the scheduler agent or SFC inform both the machine agent and the order agent
agent). The function of the shop floor control agent is requesting that task. The order agent and machine
to match the request of the order agent with the offers agent will then autonomously settle the details of the
of the resources on a certain instance of time. The task to be executed. This includes setting up the
SFC agent therefore uses the schedules it receives machine i.e. the workstation, co-ordinating their
regularly from one or more schedules. Order agents activities in time, loading the necessary auxilliary
and resources (or workstation) are not obliged to wait resources and exchanging information like NC-
with their request and offers until the operation is programs. The machine (workstation), in principle,
executable or the resource is free. This enables the has the possibility to schedule the tasks. This can be
agents to foresee future events and consider the done to account for small durations in the predicted
consequences of it. This will be used to have, e.g. an processing time; to react to resource breakdown; or to
idle machine (workstation), waiting for an important optimize setups that have not been foreseen in
order, even when work is available. sufficient detail on higher level.

In addition to the global scheduler, each machine

(workstation) is equipped with a local scheduling

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec 2017 Page: 468
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
4.0 Agent Structure Modeling uses constraint propagation computation to adjust
generated values.
In this work, the Belief - Desire – Intention (BDI)
agent model is adopted to model the system. The BDI  Data on available machines (i.e. available
software model is one of the software models resources):
developed for modeling and programming intelligent The machine agent from time to time updates the
agents [17]. scheduler agent on the availability of machines (in
In BDI terms Beliefs represent knowledge of the other words data on free machines).
world. It represents information state of the agent. In
computational terms, Beliefs represent the state of the  Request Order:
world, such as the value of a variable (i.e. constraint A data structure that list and stores request received
variable in the proposed job shop scheduling problem from the order agent.
(JSSP) model), a mathematical or logical expression
(in this design constraint inqualities, JSSP constraint  Job state space probability vector:
expressions, and scheduling constraint equations), a This is data of stochastic projection received from the
relational database and belief includes inference rules shop floor agent. The shop floor agent uses Markov
used by the agents that allow forward chaining that chain to arrive at this projection.
leads to new beliefs.
 Desire
Desire (or goals) form another essential component of The scheduler agent has the desire
the system state. Desire represents the motivational  Makespan Minimization:
state of the agent, objectives or goal that the agent The objective (i.e desired end state ) of this agent is to
would like to accomplish [17]. In computational apply it’s scheduling heuristics in order to minimize
terms, a desire may simply be the value of a variable, the manufacturing makespan.
a record structure, or a symbolic expression in some
logic. The important point is that a goal represents  Assignment of jobs (unscheduled jobs,
some end state. preempted jobs) to machines:
The agents want to arrive at the state whereby all
Intentions represent the deliberative state of the agent, unscheduled jobs, pre empted jobs, waiting jobs are
the agents chosen actions, which means the agent has completely assigned to machines.
begun executing a plan. Computationally, intentions
may simple be a set of executing threads in a process  Constraint Satisfaction:
that achieve the goals (desires) of the system. The agent objective is for all mathematical constraint,
that would be presented in detail in this chapter later
4.1. Specifying the BDI components of the are satisfied for whatever value of variable (e.g. start
proposed agent based model. time, processing time etc) are computed by the
Referring to the proposed system block diagram and scheduler algorithm as a possible solution to the JSSP.
narrative of section 3.2:
 Intention
 Scheduler Agent BDI structure The scheduler agent carries out the following action
(intention); these actions are methods or threads in
 Belief software systems. This explains why the
The agent has the following belief; representation is in form of program function
 Scheduling Constraints (Resource Constraints, (methods).
task constraints, temporary constraints):  Run scheduling Algorithm ()
This constraint, are specified in this model design in The scheduler agent runs the algorithm that schedules
form of mathematical expressions, inequalities, (i.e. assigns) jobs to machines. This algorithm is
logical expressions that the computed schedule specified in this chapter.
parameter must satisfy.To this effect the scheduler
agent, uses combinatorics (i.e a search) to generate  Probe Resource Availability ()
scheduling values (start time for jobs, and machine) The scheduler agents continuously run the code to get
that ensures that the set constraints are satisfied. It information on available resource (machines) from the

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec 2017 Page: 469
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
order agent and machine agent. This is necessary for it  Global optimization (Markov chain)
to compute the next schedule, to update the current The agent wants to keep the schedule optimized at
schedule or discard the current schedule. every time t. it uses the Markov chain to project a
forecast state space that enables the schedule agent to
 Send schedule to shop floor Agent () adopt its Runscheduling Algorithm() intentions.
This scheduler agent intention computes and sends This is a key part of the proposed system.
schedule to the shop floor agent. It is the shop floor
agent that executes the schedule.  Job and Machine (Resource) Alignment
This agent wants to actualize the objective of
 Execute Constraint Propagation () matching job(ij) with the right machine. Its Execute
This intention is specified to run verification code by Schedule () intentions is the action it carries out to
evaluating the various constraints that have been realize this objective.
specified for the JSSP. These constraints are given as
mathematical expression in this design. By evaluating  Keep Scheduler Agent Update on Job State
constraint satisfaction for any possible solution for the Transient
JSSP, the agent constructs new constraints (temporary This desire is from the perspective of Markov Chain.
constraints) and updates its belief. The agent wants to let the schedule agent to be aware
of the probability distribution of the non-absorbing
 Shop floor control Agent BDI structure states of the systems at every scheduling loop or
 Beliefs scheduling window or upon request from the
The agent has the following beliefs scheduler agent.

 Machine Status  Intentions

It gets knowledge (i.e data) on the status of machines The shop floor control agent carries out the following
(idle, breakdown etc) from the machine agent. It gets actions (operations or plan execution):
this knowledge in form of updates.
 Initiate Markov Chain ()
 Manufacturing Schedule The agent uses the job status updates from the order
It gets knowledge of completed schedules from the agent to constitute the job state space vector at time t,
scheduler agent. The schedule is the data on the it evaluates the mathematics to obtain the Markov
assignment of jobs to machines. transition matrix. It uses the transition matrix to
project the initial system state space to incorporate
 Job Status stochasticity in the schedule computation of the
This is the states of all jobs at time t. It has knowledge schedule agent. The outcome of this intention gives
of the state of all jobs in the system at any time. The the schedule agent the initial distribution (the highest
status of the job could be: Unscheduled, scheduled, probability likelihood) from which the schedule is
waiting, processing, preempted (i.e stop or computed.
interrupted), finished. It constitutes this belief as its
state space () at time t.  Execute Schedule ()
It executes the schedule from the schedule agent. The
 Manufacturing Request schedule gives this agent the required instructions on
The encoded instruction from the order agent on the the allocation of operations to machines. The
details of requested manufacturing. What sequence of execution of the schedule enables it to pass control
operations should be carried out on a job, the instructions to the machine agent.
sequence of machining required to deliver the job.
 Run Update on Allocation ()
 Order Request It updates the scheduler agent belief, the machine
A registry data structure that constitutes a listing of agent belief and the order agent belief on machine (i.e
order received. resource) allocation. It also updates the scheduler
agent on job state space at time t.
 Desires
The agent desires:

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec 2017 Page: 470
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
4.2 Modeling the Scheduling Constraints between the variables, assign a value to each variable,
so that all the constraints are satisfied. Hence the
The core of the algorithm is the agent BDI computing principle observed by the Scheduler Agent is that
of the schedule and Markov process. In a continuous given a set of resources (i.e machines) with given
and iterative process, the agent based modeling is capacities, a set of operations or activities (to be
used to compute and track the state of jobs to produce carried out at the machines) with given processing
initial state distribution which is then used to initialize times and resource, requirements, and a set of
the Markov Chain to produce the optimized target temporal constraints between activities, the
schedule. The scheduler agent uses constraints scheduling problem consists of deciding when to
programming technique to produce the schedule. execute each activity or operations O, so that both
temporal constraints and resource constraints are
Most scheduling problems can easily be represented
satisfied.
as instances of the constraints satisfaction problem
[18]: given a set of variables, a set of possible values
(domain) for each variable, and a set of constraints

 Makespan Objective Function

The scheduler agent’s desire (goals) is to:
Minimize the manufacturing completion time, or makespan (MS) for processing all jobs. Mathematically, the
problem of minimizing the manufacturing makespan is equivalent to the following formulation (adopted from
mixed integer programming) [17]

Min Ms = f (P ), Min Tmax

ijk
Subject to (i.e. the constraints):
Bij +  Pijk Xijk ≤ Bi, j +1 ¡=1,2,....N; j = 1, 2, …j(i) -1 …..... (4.1)
K є Mij

Constraint set (4.1) ensures that an operation or activity j + 1 cannot start before the previous
operation j of the same job ¡ has been completed.
Bi,J(i) +  Pi,J(i)k – MS ≤ 0, ¡ = 1, 2, …….., N ……................(4.2)

K є Mi,J(i)

constraint set (4.2) ensures that the starting time and processing time of the last operation J(i) for
job ¡, ¡ = 1, 2, ……., N is less than or equal to the makespan (MS).

 Xijk = 1, ¡ = 1, 2,……, N; j = 1, 2, ….. J(i) ….......... (4.3)

K є Mij

Equation (4-3) ensure that one operation j of job ¡ can only be performed on only one machine k at
a time. In essence, this constraint guarantees that each job ¡ takes only one path through the
system.
Xghk + Xijk -1 ≤ Yijghk + Yghijk, …………………………. (4.4)

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec 2017 Page: 471
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
¡ = 1, 2, ……….., N; g = 1, 2, …….,N: ¡‡g; j = 1,2, ……, J(i);

h = 1, 2,………..,J(g); k є Mij Ո Mgh.

During scheduling decision or search, the Agent uses constraint set (4.4) to restrict two operations of
two different jobs that are scheduled on the same machine from being performed at the same time.
Thus, only one operation of one job is always performed before the other operation of the second
job.
Yijghk + Yghijk ≤ 1,

¡ = 1, 2, ……., N ; g = 1, 2, …….. , ¡‡g; j = 1, 2, ……J(i) …….....(4.5)

h = 1, 2, …….., J(g); k є Mij Ո Mgh.

(Bij + Pijk Xijk) – (Bgh + Pghk Xghk) + Y (1-Yghijk)≥ Pijk Xijk

¡ = 1, 2, …….N, g = 1, 2, …….. N, ¡‡g; ......................... (4.6)

j = 1, 2, ……….., J(g), h = 1, 2,……….J(g); k є Mij Ո Mgh

(Bgh + Pghk Xghk) – (Bij + Pijk Xijk) +Y(1-Yijghk)≥Pghk Xghk …….(4.7)

¡ = 1, 2,………N; g = 1, 2, …….,N; ¡ ‡ g; j = 1, 2, …….J(i);

h = 1, 2, …….., J(g); k є Mij Ո Mgh

The agents of constraint set (4.5) in its belief guarantees that if operation j and h from jobs i and g,
respectively, are to be performed on the same machine k, then the two operations cannot be
performed simultaneously. Agent implements constraint set (4.6) to ensure that if operation j of job
i is chosen to be processed before operation h of job g, the starting time and processing time of
operation j of job i must be less than the starting time of operation h of job g. The same logic
applies to constraint set (4.7) for the reverse case when operation h of job g is chosen to be
processed before operation j of job i. Agent, these constraints reinforce that one job is always
processed before a second job on a given machine to avoid contacts.
Bi1 ≥ Ri, ¡ = 1, 2, …………N ………………....................... (4.8)
Constraint set (4.8) ensures that the first operation of a job i cannot start before it is ready.
Bij ≥ 0 ¡ = 1, 2,………… N; j = 2, ………. J(i) ……….......... (4.9)
MS ≥ 0 ……………….. (4.10)
Using the non-negativity constraints (4.9) and (4.10) the agent ensures that all starting times for
the remaining operations and the manufacturing makespan are positive.

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec 2017 Page: 472
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
Xijk є {0, 1}, ¡ = 1, 2, ………, N; j = 1, 2, ……..J(i); ……(4.11)
k = 1, 2, …………,m
Yijghk є {0, 1}, ¡ = 1, 2, ………….N, j = 1, 2……………J(i);
g = 1, 2, ……………..N; h = 1, 2, ……,J(g)
k = 1, 2, ……………..,m
Constraints (4.11) and (4.12) show the integer constraints for the 0-1 variables.

Bi, j(i) + X i, J(i)k = Ci, ¡ = 1, 2, . . … N … . . (4.13)

є , ()

C¡ - 0¡ - Tmax ≤ 0 ¡ = 1, 2, ……….,N …………………… (4.14)

Tmax ≥ 0 ……………………….........................(4.15)

Constraint set (4.13) ensures that the starting time and processing time of the last operation J(i) for
job ¡, ¡ = 1, 2, …….N is equivalent to the manufacturing completion time, while constraint set (4.14)
ensures that the tardiness of job ¡, ¡ = 1, 2, ……………….N is less than or equivalent to the maximum
tardiness. With constraint (4.15) the scheduler agent ensures the maximum tardiness value will be
non negative.

Iik + ∑ Xijk + 1Pijk + Bij + 1, k = Ij k + 1 + ∑ X ijk Pijk + 1 + Bj k 1 ………………. (4.16)

, , +

Constraint (4.16) established the relationships required to keep the consistency between machine’s idle time
and machine blocking times.
The scheduler agent makes scheduling decisions. This decision involves searches for time values (i.e start time
of operations i on machine Mk) that satisfy all constraints as presented.

5.1 Conclusion
When making scheduling decision, the scheduler into the solution of the JSSP in the proposed model.
agent continuously carry out constraint propagation in The methodical leverage of the contributions of this
order to determine conditions (resulting in temporary work would help remove the bottleneck currently
constraints) that a schedule must satisfy (as it relates inherent at the shop floor towards the effective
to operation ordering) to meet all the considered exploitation of these production management
constraints. This results in the scheduler activating the strategies.
intentions that updates the time-bounds of each of the
two operations. The algorithm specified in this work REFERENCES
for the scheduler agent is based on the earliest time
1. Albert Jones, Luis C. Rabelo “Survey of job shop
heuristics [19]. To enable the incorporation into the
scheduling techniques” MIT, USA (2009), pp 232
shop floor agents inference (i.e intention) the
stochastic effect of events at the shop floor, there has 2. Eric scherer “shop floor control – An integrative
to be a way to make stochastic projections from which framework from static scheduling models towards
the scheduler agents intention (Run scheduling an agile operations management” , CH8092
Algorithm()) can optimize the schedule computation Zurich Switzerland 2002, pp. 1,2.
(which is more of a combinatorics). For the scheduler 3. Philippe Baptiste, claude le pepe and Wim
and shop floor agents’ deliberations (intentions) to Nuijten. Constraint-Based optimization and
factor in stochasticity, Markov Chain is integrated

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec 2017 Page: 473
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
Approximation for job-shop scheduling” ILOG 15. Dantzig, G. and P. walfe “Decomposition
S.A., France (1995). principles for linear programs”, 8(1): 101-111
4. Baker, K.R [1974]. “Introduction to sequencing 16. Bauer, A, Bowden, R., Brown, I. Duggen J en
and scheduling, Wiley & sons. Lyons, G., “shop floor control systems, design and
implementation, “chapman & Hall, London, 1991.
5. Akeela M.Al-Atroshi, Sama T. Azez, Baydaa S.
Bhnam An Effective Genetic Algorithm for job 17. Ten Noranis Mohd Aris “Object-oriented
shop scheduling with fuzzy Degree of programming semantics representation utilizing
satisfaction,” University of Dohuk, Irag (2013), agents” journal of theoretical and applied
pp. 180 information technology sep. 2011.vol. 3, No. 1 Pp.
10, 11, 12.
6. Z. Othman, K. Subari, and N. Morad, “job shop
scheduling with Alternative machines using 18. Nuijten, W. P. M & Aarts, E. H. L. (1994)
Genetic Algorithm,” journal Tek nolosi, 41 (0). “Constraint satisfaction for multiple capacity job-
2007.pp. 67-78 shop scheduling” proc. 11th European, Conference
or Artificial Intelligence, pp 635-639. John Wiley
7. K. Raya, R. Saravaran, and V. Selladurai,” Mult-
and Sons.
Objective Optimization of parallel machine
scheduling (PMS) using Genetic Algorithm and 19. Grimm, V.,U. Berger, F. Bastiansen, S. Eliassen,
fuzzy Logic Approach” IE(I). journal-PR,vol. 87, V. Ginot, J. Giske, J. Goss-Custard, T. Grand,
2008, pp. 26-31 S.K.Hein Z,G.Huse, A Huth, J.U.Depson, C.
Jorgansen, W.M.Mooij.B. Muller, G. Pe’er, C
8. Blazewic, S. Ecker, K. H., Pesch. E., Schmodt, G.
piou, S.F. Railsback, A.M.Robbings, M.M
Ind J. Weglerz, Scheduling Computer and
Robbins, E.Rossmanth, N. Ruger, E. Strad, S.
Manufacturing processes, Berlin (springer) 2001,
Souissi, R.A. Stillman. R. Vabo, U. Visser, and
ISBN 3-540-41931-4. PP. 322, 323.
D.L. De Angelis. (2006). “A standard protocol for
9. Sven Banisch, Ricardo Lima and Tanya Araii Jo describing individual-based and Agent-Based
“Aggregation and Emergence In Agent-Based models” Ecological modeling 198(1-2): 115-126.
models: A markov chain approach” ISEG, Lisbon
(2012). ISSN 0874-4548.
10. Izquiredo, L.R., Izquiedo, S.S., Galan, J.M., and
Santos, J.I.(2009). “Techniques to understand
computer simulations: Markov chain analysis”.
Journal of Artificial societies and social
simulation, 12(1):6.
11. C. Low, C.M. Hsu and K.I. Huang, “ Lot sputting
in job shop scheduling using integer
programming”. International Journal of advanced
manufacturing technology, vol. 24, pp773-780,
2004.
12. U. Buscher and L Shen. “An integrated tabu
search algorithm for the lot streaming problem in
job shops” European journal of operational
research, vol. 19 pp 385-399, 2009.
13. Davis, W. and A. Jones (2001), ” A real-time
production scheduler for a stochastic
manufacturing environment”. 1(2): 101-112.
14. Benders, J. “Partitioning procedures for solving
mixed-variables mathematical programming
problems” Number sche mathematik, 4(3): 238-
252.

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec 2017 Page: 474