IOP Conference Series: Earth and Environmental Science

PAPER • OPEN ACCESS

An Integer Programming Model for Flow Shop Scheduling Under TOU

and Tiered Electricity Price
To cite this article: Lining Peng et al 2021 IOP Conf. Ser.: Earth Environ. Sci. 692 022105

View the article online for updates and enhancements.

This content was downloaded from IP address 196.188.115.240 on 31/08/2021 at 09:04

EMCEME 2020 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 692 (2021) 022105 doi:10.1088/1755-1315/692/2/022105

An Integer Programming Model for Flow Shop Scheduling

Under TOU and Tiered Electricity Price

Lining Peng*, Mingshun Yang and Renhao Xiao

School of Mechanical and Precision Instrument Engineering, Xi’an University of
Technology, Shaanxi, China

*Corresponding author: [email protected]

Abstract. Reasonable scheduling can achieve energy saving in the production process.
Considering the TOU and tiered electricity price policy, an integer linear
programming model for flow shop with the goal of minimizing the power cost is
constructed, and a particle swarm optimization algorithm (PSO) with a three-layer
coding method based on the workpiece, start time and machining state and the
iterative process is designed. Finally, the correctness of the mathematical model and
the effectiveness of the algorithm are verified by an example.

Keywords: Flow shop, TOU and tiered electricity price, PSO, production scheduling.

1. Introduction
According to the National Bureau of Statistics, the industrial sector accounts for 70 percent of the total
energy consumption, 83.5 percent of which comes from manufacturing. As a direct secondary energy,
electric energy is widely used in the manufacturing industry. Demand Side Management (DSM) is to
guide power users to change their electricity consumption mode by adopting incentive measures (such
as time-of-use price policy), so as to truly realize the conservation and efficient use of energy [1].
Wu [2] studied the same time parallel machine scheduling rate and a continuous-time mixed integer
linear programming (MILP) model was established to minimize the total power consumption. Ding [3]
studied hybrid flow shop scheduling problems under time rate (HFSP), the purpose was to minimize
the total electricity in order to ensure the completion of tasks in the production process during the
program. Zhang et al. [4] studied the scheduling problem under time-of-use electricity price to
maximize the completion time and electricity. For the mixed flow shop scheduling problem, Wang et
al. [5] proposed a mixed integer programming model to minimize the total energy consumption and
completion time.
Time-of-use electricity price is divided into several periods on the basis of unified electricity price.
Different electricity prices are adopted for different periods, generally divided into peak, flat and
trough periods. A single time-of-use electricity price may cause the shift of peak and valley periods,
and a single step price cannot achieve the purpose of grid stability. Therefore, the TOU price is
proposed [6] which combines the TOU price and the step price, to achieve the purpose of energy
saving and emission reduction. This paper adopts discrete-time modeling under the consideration of
TOU and tiered electricity price, takes the minimum electricity cost as the goal, establishes a

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
EMCEME 2020 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 692 (2021) 022105 doi:10.1088/1755-1315/692/2/022105

mathematical model for the production scheduling problem, and proposes a particle swarm algorithm
to solve the model. Finally, an example is given to verify the established model and algorithm.

2. Problem description
The flow shop scheduling problem considering TOU and electricity price stratification is described as
follows. Each workpiece must be processed by multiple machines with the same processing path, one
machine can only execute one process. TOU and hierarchical electricity prices are based on the user's
electricity demand curve, which divides the processing period into three types: peak period, flat peak
period and valley period. A multi-level price for each segment is predetermined.
Given the processing time and power requirements of each workpiece on the corresponding
processing machine, considering the time-sharing step price, the processing sequence and start
processing time of each workpiece are required to ensure that the processing task is completed within
the specified time and minimize processing Capacity cost.
Following assumptions are considered: ① Each machine can only handle one process at a certain
time. ② Processing time is known and predetermined. ③Once each process starts, the process cannot
be interrupted. ④ Processing time of each process includes the preparation time. ⑤ Each process of
the workpiece must be completed before subsequent processes.

3. Model building
In order to establish a scheduling model, the mathematical symbols are defined as follows:
M: Manufacturing machine, i  1, 2,3,..., M ; N: Production target, j  1, 2,3,..., N ;
T: Specified duration, t=1, 2, 3, …, T;
PTij: The processing time of workpiece j on the machine i;
qij: Machine i power requirements for workpiece j;
Nit: The total number of products completed on machine i at time t;
xijt: Machine i processes workpiece j at time t, then xijt=1, otherwise xijt =0;
yijt : Machine i processes workpiece j at time t, then yijt =1, otherwise yijt =0;
S: The number of sub-periods, s  1, 2,3,..., S ;
V: The number of electricity price steps, v  1, 2,3,...,V ;
Rsv: The electricity consumption of period s is included in the range of the vth gear price, then Rsv=1,
otherwise Rsv =0;
PCSsv: The dividing point of the v-th step power consumption in time period s;
Psv: The v th price of period s, Electricity consumption in period s;
TEC: Total power cost.
Without considering electricity consumption such as lighting, only considering power consumption
of production machines, the mathematical model is established as follows:

min TEC (1)

Among them:
 
 
S V
TEC    Ps1 Es   Rsv  Es  PCSsv   Psv  Ps ,v 1   (2)
s 1  v 2 

M N
Es   xijt qij (3)
ts i 1 j 1

In view of the research problem, consider the following constraints:

N T
Nit   yijt , NiT  N (4)
j 1 t 1

2
EMCEME 2020 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 692 (2021) 022105 doi:10.1088/1755-1315/692/2/022105

k  PTij

t k
xijt  yijk PTij 1 k  T (5)

M N

x
i 1
ijt  1, x j 1
ijt 1 (6)

Nit  Z , xijt , yijt , Rsv  {0,1} (7)

Among them: Formula (1) represents the objective of minimizing the total power cost ; Formula (2)
represents the total power cost in the processing ; Formula (3) represents the electricity consumption
of each period; the constraints (4) ensures that the processing task is completed within the specified
time, and each workpiece completes the specified number of procedures; the condition (5) guarantees
that the machining process is not interrupted, and once the process of the workpiece starts processing,
it cannot be interrupted; the constraint condition (6) is to ensure that a machine can only process one
workpiece at the same time, to meet the machine constraints; constraint (7) ensure that a workpiece
can only be processed by one machine at the same time to satisfy the workpiece constraints.

4. Algorithm design
The flow shop scheduling that considers TOU and tiered electricity price increases the complexity of
the problem further. PSO was proposed in 1995, and more and more scholars are improving it to solve
the NP-Hard problem. In this paper, the particle swarm algorithm is used to solve the problem, and
according to the complexity of the problem, redesign the key part of particle swarm.

4.1. Coding method

The flow shop scheduling problem under the TOU and tiered electricity price requires comprehensive
consideration of the processing sequence and processing time of the workpiece. In order to facilitate
the determination of the start processing time of the workpiece and the time period of machine
processing, three layers of coding are adopted: workpiece code + starting time code + processing
status code, the first layer is the workpiece code, which is encoded based on the processing sequence
of the workpiece. The second layer is the starting time code, which is based on the starting processing
time of each machine after discretization of the specified time. The third layer is the machine
processing state code, which is based on the processing state of the machine.

4.2. Location updates operation

For the workpiece code, the precedence preserving order-based crossover (POX) is used. For the start
time code and the processing status code, the position update is also carried out according to the
probability. Two points are used to cross, and then two points are generated. The part between two
points is exchanged, and the illegal solution will be generated after the position update. A correction
method is used to ensure the feasibility of the solution.

4.3. Position change operation

For each individual, the position transformation in probability for operation, using two exchange ways,
the corresponding coding layer randomly generated two locations, is located in the two positions
coding swapping. For the workpiece code, this position transformation operation won't produce illegal
solution for the processing of time code and status code, after the illegal solutions are generated, a
later correction method is used to ensure the generation of feasible solutions.

4.4. Algorithm flowchart

The algorithm is divided into two parts, the flowchart is shown in Figure 1. The first part is to perform
position update and position conversion operations on the workpiece code. When the workpiece can

3
EMCEME 2020 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 692 (2021) 022105 doi:10.1088/1755-1315/692/2/022105

complete the processing task within the specified time, the second part is executed, and the time code
and the status code perform position update and position transformation operations. The second part
has a larger search range, so the second part performs a search every time. When the number of
iterations of the entire population meets times, the optimal solution in the iteration process is output.

5. Case analysis
For the flow production workshop of a machine tool manufacturing enterprise considering the TOU
and tiered electricity price, the shaft parts are processed. There are 4 machines in the workshop, and
there are 2 different shaft workpieces, 5 pieces of each type, and a total of 10 pieces of artifacts need
to processing, it is required to complete the processing task within 6 hours. The processing time and
power requirements of each type of workpiece are shown in Table 1.

Figure 1. Algorithm flowchart.

Table 1. Processing Time and Power of Each Machine

Workpiece (process time (min), power demand (KW)）
Category Machine1 Machine2 Machine3 Machine4
1 (25,15) (20,20) (30,25) (25,20)
2 (20,20) (25,20) (20,25) (25,20)

Table 2. The TOU and Tiered Electricity Price Period and Price List
Peak period Flat peak period Trough period
First gear 0.9831 0.6712 0.3594
Second gear 1.3131 1.0012 0.6894

With reference to the current time-of-use electricity price and tiered electricity price, set two levels
of electricity prices for each time period, and construct the TOU and tiered electricity price model of
this article with a specified time of 6 hours. In order to reduce the complexity of the problem, we take
12 per unit times as an electricity price time. There are 6 electricity price periods within the specified
time period, including 2 trough periods, 3 flat peak periods and 1 peak period. Assuming that the

4
EMCEME 2020 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 692 (2021) 022105 doi:10.1088/1755-1315/692/2/022105

electricity price per period is shown in Table 2. The cutoff point of the two-tier price of the TOU and
tiered electricity price is 45 KWH/h.
According to the model established above, the algorithm was written in the m language, and the
program was run in MATLAB R2014b. The population size is 50,  =0.5, Pm=0.1, D=500, DT=800.
After running 10 times, the optimal value was selected, and the scheduling Gantt chart was obtained,
as shown in Figure 2.
Different from the above-mentioned goal of minimizing power cost, the comparison group aims to
minimize the maximum completion time, and the program runs in the same environment. After 10
times of execution, the optimal value is obtained, and the scheduling Gantt chart is shown in Figure 3.
It can be seen from Figure 2 that the scheduling considering the TOU and tiered electricity price
effectively completed the processing task within the specified time, and the starting time of the
workpiece was reasonably arranged within the specified time. From the comparison of Figure 2 and
Figure 3, the electricity cost comparison of the two groups can be obtained. It can be seen that the
dispatch results obtained in this paper considering the TOU and tiered electricity price have effectively
reduced the electricity cost by 11%.

Figure 2. Scheduling Gantt diagram

Figure 3. Gantt diagram of comparison group scheduling

5
EMCEME 2020 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 692 (2021) 022105 doi:10.1088/1755-1315/692/2/022105

6. Conclusions
In this paper, a mathematical model of the flow shop scheduling problem considering the hierarchical
electricity price is established and the corresponding PSO algorithm is designed. Finally, an example
is given to verify the correctness of the model and the effectiveness of the algorithm. The results show
that scheduling considering time-of-use electricity price and electricity price stratification can reduce
the production cost of enterprises to a certain extent and has certain guiding significance for
production. This article does not consider the waiting cost of the production process, such as the
standby energy consumption of the machine.

Acknowledgments
This work is supported by National Natural Science Foundation of China (Grant No. 52005404),
China Postdoctoral Science Foundation (Grant No: 2020M673612XB) and Doctoral innovation fund
of Xi'an University of Technology (Grant No: 310-252072013).

References
[1] Gahm C, Denz F, Dirr M, et al. Energy-efficient scheduling in manufacturing companies: A
review and research framework [J]. European Journal of Operational Research, 2016, 248 (3):
744–757.
[2] Wu Y J. Hybrid Tabu Search Algorithm for Parallel-Machine Scheduling Optimization under
Time-of-Use Tariffs [J]. Modern Computer, 2019 (13): 26–31
[3] Ding J Y, Song S J, Zhang R, et al. Parallel Machine Scheduling Under Time-of-Use Electricity
Prices: New Models and Optimization Approaches[J]. IEEE Transactions on Automation
Science and Engineering, 2016, 13 (2): 1138–1154.
[4] Zhang M Y, Yan J H, Zhang Y L, et al. Optimization for energy-efficient flexible flow shop
scheduling under time of use electricity tariffs [J]. Procedia CIRP, 2019, 80: 251–256.
[5] Wang S J, Wang X D, Chu F, et al. An energy-efficient two-stage hybrid flow shop scheduling
problem in a glass production [J]. International Journal of Production Research, 2020, 58 (8):
2283-2314.
[6] Liu B, Jiang B L, Hao N, et al. Coordinated optimization of TOU & tiered pricing and optimal
scheduling model combining microgrid under uncertain bilevel programming [J]. Power
System Protection and Control, 2019, 47 (07): 75–83.