Source-linked AI summary
Coordinating Flexible Demand Response and Renewable Uncertainties for Scheduling of Community Integrated Energy Systems with an Electric Vehicle Charging Station: A Bi-level Approach
Yang Li, Meng Han, Zhen Yang, Guoqing Li
TL;DR
The paper addresses coordinated CIES–EVCS scheduling under flexible demand and renewable uncertainty in a multi-stakeholder setting. It proposes a bi-level model with integrated demand response and TOU-plus-real-time pricing, and reports balanced stakeholder interests and improved joint operation through coordinated EV and renewable scheduling.
Problem
Coordinating CIES and EVCS dispatch is challenging because renewable volatility and disorderly EV charging complicate scheduling and joint-cost reduction.
Method
A bi-level CIES–EVCS model combines integrated demand response, TOU-plus-real-time pricing, EV charging, discharging, and spinning-reserve participation.
Results
The method balances CIES and EVCS interests, maintains user satisfaction within an acceptable range, reduces joint operation costs, and improves economic performance through EV spinning reserves.
Takeaways & Limitations
Coordinating flexible demand, EV behaviors, spinning reserves, and renewable uncertainties provides an effective dispatching strategy for the studied practical CIES.
Abstract
from arXiv · showhide
A community integrated energy system (CIES) with an electric vehicle charging station (EVCS) provides a new way for tackling growing concerns of energy efficiency and environmental pollution, it is a critical task to coordinate flexible demand response and multiple renewable uncertainties. To this end, a novel bi-level optimal dispatching model for the CIES with an EVCS in multi-stakeholder scenarios is established in this paper. In this model, an integrated demand response program is designed to promote a balance between energy supply and demand while maintaining a user comprehensive satisfaction within an acceptable range. To further tap the potential of demand response through flexibly guiding users' energy consumption and electric vehicles' behaviors (charging, discharging and providing spinning reserves), a dynamic pricing mechanism combining time-of-use and real-time pricing is put forward. In the solution phase, by using sequence operation theory (SOT), the original chance-constrained programming (CCP) model is converted into a readily solvable mixed-integer linear programming (MILP) formulation and finally solved by CPLEX solver. The simulation results on a practical CIES located in North China demonstrate that the presented method manages to balance the interests between CIES and EVCS via the coordination of flexible demand response and uncertain renewables.
I. INTRODUCTION
The paper addresses coordinated scheduling of CIES and EVCS under renewable uncertainty by combining flexible demand response, EV participation, and multi-stakeholder optimization. It also models flexible electrical and thermal loads while constraining user comfort and satisfaction.
- Renewable volatility causes curtailment and complicates CIES scheduling, while disorderly EV charging further increases EVCS and system-load fluctuations.
- Prior studies coordinate energy carriers, storage, demand response, or EV charging, but generally omit a comprehensive multi-stakeholder CIES–EVCS dispatch.
- The proposed model explores EV charging, discharging, and spinning-reserve provision while coordinating flexible demand responses and renewable-generation uncertainties.
- Flexible electrical demand includes shiftable and interruptible loads, with shiftable consumption time adjustable while total electricity consumption remains constant.
- Interruptible electrical load can be reduced during insufficient supply or high prices, with its maximum interruption set to 10% of period demand.
- Heating demand response links building temperature and heat demand, using PMV to represent acceptable thermal comfort and user comprehensive satisfaction to evaluate IDR effects.
C. Electric Vehicle Charging Station
The EVCS model represents uncertain vehicle arrivals, departures, travel mileage, charging demand, and state of charge to construct an initial disordered charging scheme.
- EV arrival and departure times are modeled with probability distributions whose means and standard deviations describe connection-time uncertainty.
- Daily EV travel mileage is modeled probabilistically, with its distribution parameterized by mean mileage and standard deviation.
- Actual end-of-charge SOC is determined from travel mileage, initial SOC, per-100-kilometer energy demand, and battery capacity.
- Charging time depends on required charging state, rated EV power, charging efficiency, and battery capacity.
- Monte Carlo simulation converts arrival, travel, and charging uncertainties into an EV disordered charging demand used as the first-iteration EVCS scheme.
III. PROBLEM FORMULATION
The problem formulation uses a bi-level CIES–EVCS dispatch in which operating costs are minimized at both levels and dynamic pricing coordinates their decisions. The pricing process combines renewable expectations, TOU prices, and lower-level EV responses.
- III. PROBLEM FORMULATION: The upper and lower levels minimize CIES and EVCS operating costs, respectively, with “TOU+RT” pricing linking their decisions.
- A. Dynamic Pricing Mechanism: The dynamic pricing mechanism combines TOU and RT pricing to guide lower-level EV charging-discharging decisions and reduce both operating costs.
- A. Dynamic Pricing Mechanism: The pricing relationship uses CIES load demand, expected renewable output, ESS, heat-storage, and EV charge-discharge powers.
- A. Dynamic Pricing Mechanism: Dynamic prices passed from CIES to EVCS are defined relative to grid TOU prices and valley-period TOU prices.
- A. Dynamic Pricing Mechanism: Price concessions let users select lower-price periods, reducing electricity costs and encouraging participation in integrated demand response.
- A. Dynamic Pricing Mechanism: The pricing process first obtains expected renewable output using SOT, then optimizes lower-level EV behavior and feeds it back to the upper level.
B. The Upper-level Model
The upper-level model minimizes CIES net operating cost subject to power, storage, reserve, and heating-system constraints. It accounts for grid, ESS, EV, and thermal-system operating variables.
- The upper-level objective is minimization of the CIES net operating cost.
- Operating costs include grid, ESS, and EV reserve prices, ESS depreciation, and compensation prices for interrupted electric and heating loads.
- Power-supply constraints enforce CIES supply-demand balance and limit power provided by the grid.
- ESS constraints limit capacity and charge-discharge power while requiring the same initial state at the end of each cycle.
- The ESS reserve capacity is constrained, and total spinning reserve is supplied by the grid, ESS, and EVs under a chance constraint.
- The spinning-reserve chance constraint uses α as the preset confidence level.
- Heating-system constraints enforce electrical and heat-power balances and characterize electric-boiler performance through its heating-to-electricity ratio.
- Heat-storage power and capacity constraints are analogous to the ESS constraints.
C. The Lower-level Model
The lower-level EVCS model minimizes net operating cost while coordinating EV charging, discharging, and spinning-reserve provision under operational constraints.
- EV users can lower charging costs through discharging while alleviating CIES power-supply pressure.
- The lower-level objective minimizes the EVCS net operating cost.
- EVs power balance constraints: EV charging and discharging must remain within allowable ranges while maintaining system power balance.Discharge power cannot exceed the deficits of electric and heat loads after renewable generation.
- Spinning-reserve capacity from EVs cannot exceed the grid-provided reserve requirement or the EVs’ available capacity.An additional capacity variable represents upper-level requirements when ESS reserve capacity is insufficient.
- EV power purchased from the grid or CIES is bounded by grid limits, controllable load power, and surplus renewable generation.
- EVs battery constraints: Battery constraints bound EV storage capacity and the numbers of charging and discharging vehicles.
A. Probabilistic Serialization Description of RGs
Renewable-generation uncertainties are represented as probabilistic sequences, discretized from output distributions, and combined to support chance-constraint reformulation.
- Wind-turbine output is discretized from its probability density function into a finite probabilistic sequence.Each state has an output value and corresponding probability.
- The discretization step q trades calculation efficiency against fidelity to the actual renewable probability distribution.A sensitivity analysis selects q to balance CIES reliability and economy.
- Joint renewable-generation output sequences are obtained by addition-type convolution of the individual renewable sequences.
- A binary variable W_ue,t indicates whether total reserve capacity reaches the margin between joint renewable output and its expected value.It is 1 when the reserve is sufficient and 0 otherwise.
- Replacing the chance constraint with sequence-based binary conditions transforms the CCP into a MILP formulation.
C. Determination of Dispatching Scheme
The dispatching procedure alternates upper-level CIES and lower-level EVCS optimization, passing dynamic prices and EV charging-discharging decisions between levels until termination.
- The joint optimization objective represents the operating costs of the upper-level CIES and lower-level EVCS.
- The upper level establishes the CIES model, converts chance constraints, checks feasibility, and obtains dispatch and dynamic prices.
- The lower level constructs and solves the EVCS dispatch model using prices supplied by the upper level.
- The EVCS charge-discharge scheme is used to calculate the joint objective and is fed back to the upper level during iteration.
- The procedure terminates when the iteration count exceeds the preset maximum, after which the joint optimal solution and both dispatching schemes are output.
V. CASE STUDY
The case study evaluates the proposed CIES–EVCS scheduling method using a practical winter system in North China. Results show that dynamic pricing coordinates electricity, heat, EV, and renewable operation while reducing curtailment and operating costs.
- Case setup: The practical winter case comprises WT, PV, ESS, HSD, EB, charging piles, and EVs in a North China CIES.The system includes one charging station, ten charging piles, and fifteen EVs.
- Case setup: The case study models WT and PV outputs alongside electricity and heat demands across different periods.Heat demand is determined from outdoor temperature while indoor temperature is maintained at 20°C.
- Electricity-demand scheduling: Dynamic pricing significantly reduces controllable electrical load, with EVs mainly accommodating renewable generation and reducing wind and solar curtailment.The pricing mechanism is described as beneficial for renewable accommodation.
- Heat-demand scheduling: The TOU+RT mechanism reduces grid accommodation of heat load during 1:00–5:00 while increasing renewable accommodation and maintaining heat supply-demand balance.HSD stores and releases heat energy, reducing CIES operating expenses.
D. Comparison of EVCS Scheduling Schemes under Different
The comparison shows that dynamic pricing and EV spinning reserves improve coordinated operation across the CIES and EVCS. Flexible loads further reduce peak-period purchases and system operating costs while maintaining user comfort.
- EVCS scheduling: Dynamic pricing enables EVs to participate more actively than TOU pricing and aligns charging with renewable-energy accommodation.It also guides charging and discharging to avoid centralized off-peak charging, reducing grid purchases and increasing renewable accommodation.
- EVCS economics: Dynamic pricing guides EVs toward flexible charge and discharge periods, reducing total EVCS charging-discharging costs.The comparison covers different pricing mechanisms in the EVCS cost analysis.
- EVCS economics: EV spinning reserves significantly lower EVCS operating costs during most periods compared with operation without reserve provision.Reserve participation also makes EVs more active in EVCS operation and improves EVCS economy.
- Joint operation costs: At 90% confidence, the proposed pricing mechanism reduces CIES, EVCS, and joint operating costs relative to TOU pricing.EV-provided spinning reserves additionally bring EVCS profits and reduce CIES reserve-purchasing costs.
- Iterative solution: The proposed pricing mechanism reaches its joint optimal solution in the third iteration, whereas TOU pricing reaches it in the first.The comparison is based on the iterative solving process shown in Fig. 10.
- Flexible demand response: Shiftable load decreases during peak prices and increases during valley prices, reducing peak-period grid purchases and total CIES operating cost.Interruptible electrical and heat loads operate during peak prices, achieving peak shaving within an acceptable thermal comfort range.
H. ESS and HSD Scheduling Schemes Analysis
The analysis examines ESS and HSD scheduling, confidence-level effects on costs and reserve capacity, SOT step-size sensitivity, and comparison with HIA. Results indicate coordinated storage operation, a 4–5 kW step-size range, and advantages over HIA in cost and calculation time.
- ESS and HSD Scheduling Schemes Analysis: ESS and HSD charge or store energy during valley periods and discharge or release energy during peak periods, supporting peak shaving and valley filling.This operation alleviates the system’s power-supply pressure during peak-load periods.
- Impact of Confidence Levels on Operating Costs: Higher confidence levels increase both CIES and joint operating costs because additional reserve capacity is required to balance supply and demand.The results therefore connect reliability requirements with higher operating costs.
- Impact of Confidence Levels on Reserve Capacity: Increasing the confidence level also increases the reserve capacity required by the CIES, making a reasonable confidence level important for balancing economy and reliability.Reserve-capacity growth is identified as the direct reason for higher CIES operating costs.
- Impact of Discrete Steps on Joint Optimal Costs: A step size between 4 kW and 5 kW provides the selected compromise between optimization reliability, economic results, and computation time.Steps above 5 kW create larger optimal-cost gaps, while steps below 4 kW sharply increase computation time.
- Comparative Analysis with Other Algorithms: The proposed method outperforms HIA under various confidence levels in operating costs and requires significantly less calculation time.The comparison uses average results from 20 independent HIA runs with α=90% and q=10kW.
- Conclusion: The bi-level model coordinates flexible demand response and renewable-generation uncertainties to balance the interests of CIES and EVCS.Its integrated demand response program maintains user comprehensive satisfaction within an acceptable range while balancing supply and demand.