Energy flow optimization in multi-energy system based on spatiotemporal network flows

A multi-energy system, comprising:

• M energy storage devices, photovoltaic power generators, N micro gas turbines (MGTs), and a computing device for energy flow optimization in the multi-energy system based on spatiotemporal network flows, wherein the computing device comprises a processor and a memory for storing instructions executable by the processor;
• wherein when the instructions are performed by the processor, the processor is caused to:
• obtain an objective function of the network flows in the multi-energy system, wherein the objective function is expressed by:
• min
• ⁢
• ∑
• t
• =
• 1
• T
• ∑
• n
• =
• 1
• N
• C
• n
• ,
• t
• P
• +
• ∑
• t
• =
• 1
• T
• C
• t
• λ
• ,
• where
• C
• n
• ,
• t
• P
• represents a power generation cost of the MGTs per sub-time t within time T,
• C
• t
• λ
• represents an electricity purchase cost per sub-time t within time T, λ represents an electricity price, and P represents a power output;
• determine constraints of the network flows, wherein the constraints include a power balance constraint, an energy storage period constraint, an energy storage capacity constraint, an energy storage charging and discharging constraint, a generator power constraint, and an energy storage charge quantity constraint;
• establish a network flow model with the objective function and constraints;
• and
• solve the network flow model with a shortest-path and max-flow algorithm, to obtain an optimal power flow (OPF) in the multi-energy system;
• wherein parameters of the energy storage devices, the photovoltaic power generators and the MGTs are adjusted based on the OPF, for operating the multi-energy system;
• wherein the constraints are expressed by:
• ∑
• n
• =
• 1
• N
• P
• n
• ,
• t
• G
• +
• P
• t
• λ
• +
• P
• t
• PV
• +
• ∑
• m
• =
• 1
• M
• P
• m
• ,
• t
• dis
• -
• ∑
• m
• =
• 1
• M
• P
• m
• ,
• t
• cha
• =
• P
• t
• load
• e
• m
• ,
• t
• +
• 1
• =
• e
• m
• ,
• t
• +
• P
• m
• ,
• t
• cha
• -
• P
• m
• ,
• t
• dis
• e
• m
• ,
• 0
• =
• e
• m
• ,
• T
• 0
• ≤
• P
• m
• ,
• t
• dis
• ,
• P
• m
• ,
• t
• cha
• ≤
• P
• m
• ,
• max
• P
• n
• ,
• min
• G
• ≤
• P
• n
• ,
• t
• G
• ≤
• P
• n
• ,
• max
• G
• e
• m
• ,
• min
• ≤
• e
• m
• ,
• t
• ≤
• e
• m
• ,
• max
• ,
• where
• P
• t
• PV
• is a photovoltaic power output emitted by the photovoltaic power generators at t,
• P
• m
• ,
• t
• dis
• is a discharging power of the m-th energy storage device at t,
• P
• m
• ,
• t
• cha
• is a charging power of the m-th energy storage device at t,
• P
• t
• load
• is a load of the multi-energy system at t, em,t is a charge quantity of the m-th energy storage device at t, Pm,max is an upper limit of the charging and discharging power of the m-th energy storage device,
• P
• n
• ,
• max
• G
• is an upper limit of a power output of the n-th MGT,
• P
• n
• ,
• min
• G
• is a lower limit of the power output of the n-th MGT, em,max is an upper limit of the charge quantity of the m-th energy storage device, em,min is a lower limit of the charge quantity of the m-th energy storage device, em,0 is an initial charge quantity of the m-th energy storage device, and em,T is a charge quantity at T of the m-th energy storage device;
• wherein the processor is further caused to:
• establish a graph for the network flow model, wherein the graph includes M energy storage points B and a load point L in each period, a first auxiliary point S, a second auxiliary point E, a bidirectional arc connecting L and B, a power arc from S to L, a load arc from L to E, an energy storage arc between S and B, and an electricity storage arc BB;
• and
• determine the shortest-path by searching the arcs in the graph with the shortest-path and max-flow algorithm;
• wherein the processor is caused to obtain the OPF in the multi-energy system by acts of:
• S1, initializing each arc in the graph to enable an initial flow of each arc is 0;
• S2, for an arc with a flow smaller than a flow lower limit, setting a cost of the arc to be infinite negative value, and for an arc with a flow equal to a flow upper limit, setting a cost of the arc to be an infinite positive value;
• S3, calculating an increase margin of each arc on the shortest path, wherein the increase margin of each arc is equal to a difference value between a flow upper limit and a current flow, and selecting a min-margin with the minimum difference value;
• S4, adding a first flow corresponding to the min-margin to each arc on the shortest path, and subtracting the first flow from the current flow of the bidirectional arc;
• and
• S5, outputting the OPF in response to determining that the load after adding or subtracting the first flow is consistent with an initial load, and repeating acts at S2-S5 in response to determining that the load after adding or subtracting the first flow is not consistent with the initial load;
• wherein the processor is caused to perform the shortest-path and max-flow algorithm by acts of:
• establishing an N*1 matrix D for storing arc costs, an N*1 matrix for storing a parent arc of each arc in a path, an N*1 matrix Lab for storing a marking state of each arc, where −1 indicates the arc is not marked and 0 indicates the arc is temporarily marked;
• initializing values in the matrix P and values in the matrix Lab to be −1, and initializing the matrix D to give a maximum value;
• retrieving arcs directly connected to a start arc, wherein an ending point of the start arc is S;
• for the retrieved arcs, enabling values in the matrix D to be cost values of the retrieved arcs, changing values in the matrix Lab to be 0, recording numbers and costs of the retrieved arcs into a temporary matrix H;
• in response to the temporary matrix H being not empty, selecting a min-arc with the minimum cost from the temporary matrix H;
• in response to the min-arc arc being a set termination arc for searching, determining a path from the start arc to the min-arc arc as the shortest-path;
• and
• in response to the min-arc arc being not the set termination arc, retrieving an arc i meeting min+cost (i)<D (i) from arcs connected to the min-arc, determining a path from the start arc to a parent arc of the arc i as the shortest-path.

A non-transitory computer readable storage medium stored with computer instructions, wherein when the computer instructions are executed by a computer, the computer is caused to perform a method for energy flow optimization in a multi-energy system based on spatiotemporal network flows, the method comprising:

• obtaining an objective function of the network flows in the multi-energy system, wherein the multi-energy system includes M energy storage devices, photovoltaic power generators and N micro gas turbines (MGTs), and the objective function is expressed by:
• min
• ⁢
• ∑
• t
• =
• 1
• T
• ∑
• n
• =
• 1
• N
• C
• n
• ,
• t
• P
• +
• ∑
• t
• =
• 1
• T
• C
• t
• λ
• ,
• where
• C
• n
• ,
• t
• P
• represents a power generation cost of the MGTs per sub-time t within time T,
• C
• t
• λ
• represents an electricity purchase cost per sub-time t within time T, λ represents an electricity price, and P represents a power output;
• determining constraints of the network flows, wherein the constraints include a power balance constraint, an energy storage period constraint, an energy storage capacity constraint, an energy storage charging and discharging constraint, a generator power constraint, and an energy storage charge quantity constraint,
• wherein the constraints are expressed by:
• ∑
• n
• =
• 1
• N
• P
• n
• ,
• t
• G
• +
• P
• t
• λ
• +
• P
• t
• PV
• +
• ∑
• m
• =
• 1
• M
• P
• m
• ,
• t
• dis
• -
• ∑
• m
• =
• 1
• M
• P
• m
• ,
• t
• cha
• =
• P
• t
• load
• e
• m
• ,
• t
• +
• 1
• =
• e
• m
• ,
• t
• +
• P
• m
• ,
• t
• cha
• -
• P
• m
• ,
• t
• dis
• e
• m
• ,
• 0
• =
• e
• m
• ,
• T
• 0
• ≤
• P
• m
• ,
• t
• dis
• ,
• P
• m
• ,
• t
• cha
• ≤
• P
• m
• ,
• max
• P
• n
• ,
• min
• G
• ≤
• P
• n
• ,
• t
• G
• ≤
• P
• n
• ,
• max
• G
• e
• m
• ,
• min
• ≤
• e
• m
• ,
• t
• ≤
• e
• m
• ,
• max
• ,
• where
• P
• t
• PV
• is a photovoltaic power output emitted by the photovoltaic power generators at t,
• P
• m
• ,
• t
• dis
• is a discharging power of the m-th energy storage device at t,
• P
• m
• ,
• t
• cha
• is a charging power of the m-th energy storage device at t,
• P
• t
• load
• is a load of the multi-energy system at t, em,t is a charge quantity of the m-th energy storage device at t, Pm,max is an upper limit of the charging and discharging power of the m-th energy storage device,
• P
• n
• ,
• max
• G
• is an upper limit of a power output of the n-th MGT,
• P
• n
• ,
• min
• G
• is a lower limit or the power output of the n-th MGT, em,max is an upper limit of the charge quantity of the m-th energy storage device, em,min is a lower limit of the charge quantity of the m-th energy storage device, em,0 is an initial charge quantity of the m-th energy storage device, and em,T is a charge quantity at T of the m-th energy storage device;
• establishing a network flow model with the objective function and constraints;
• solving the network flow model with a shortest-path and max-flow algorithm, to obtain an optimal power flow (OPF) in the multi-energy system;
• operating the multi-energy system by adjusting parameters of the energy storage devices and the MGTs based on the OPF;
• establishing a graph for the network flow model, wherein the graph includes M energy storage points B and a load point L in each period, a first auxiliary point S, a second auxiliary point E, a bidirectional arc connecting L and B, a power arc from S to L, a load arc from L to E, an energy storage arc between S and B, and an electricity storage arc BB;
• and
• determining the shortest-path by searching the arcs in the graph with the shortest-path and max-flow algorithm;
• wherein obtaining the OPF in the multi-energy system comprises:
• S1, initializing each arc in the graph to enable an initial flow of each arc is 0;
• S2, for an arc with a flow smaller than a flow lower limit, setting a cost of the arc to be infinite negative value, and for an arc with a flow equal to a flow upper limit, setting a cost of the arc to be an infinite positive value;
• S3, calculating an increase margin of each arc on the shortest path, wherein the increase margin of each arc is equal to a difference value between a flow upper limit and a current flow, and selecting a min-margin with the minimum difference value;
• S4, adding a first flow corresponding to the min-margin to each arc on the shortest path, and subtracting the first flow from the current flow of the bidirectional arc;
• and
• S5, outputting the OPF in response to determining that the load after adding or subtracting the first flow is consistent with an initial load, and repeating acts at S2-S5 in response to determining that the load after adding or subtracting the first flow is not consistent with the initial load;
• wherein the shortest-path and max-flow algorithm is performed by acts of:
• establishing an N*1 matrix D for storing arc costs, an N*1 matrix for storing a parent arc of each arc in a path, an N*1 matrix Lab for storing a marking state of each arc, where −1 indicates the arc is not marked and 0 indicates the arc is temporarily marked;
• initializing values in the matrix P and values in the matrix Lab to be −1, and initializing the matrix D to give a maximum value;
• retrieving arcs directly connected to a start arc, wherein an ending point of the start arc is S;
• for the retrieved arcs, enabling values in the matrix D to be cost values of the retrieved arcs, changing values in the matrix Lab to be 0, recording numbers and costs of the retrieved arcs into a temporary matrix H;
• in response to the temporary matrix H being not empty, selecting a min-arc with the minimum cost from the temporary matrix H;
• in response to the min-arc arc being a set termination arc for searching, determining a path from the start arc to the min-arc arc as the shortest-path;
• and
• in response to the min-arc arc being not the set termination arc, retrieving an arc i meeting min+cost (i)<D (i) from arcs connected to the min-arc, determining a path from the start arc to a parent arc of the arc i as the shortest-path.