Blockchain and Computational Intelligence Inspired Incentive-Compatible Demand Response in Internet of Electric Vehicles
Abstract
ev作为基础设施,实现智慧城市的电力调度交易
-
提到分布式但其实并不
propose a consortium blockchain-enabled secure energy trading framework for electric vehicles with moderate cost.
-
这里区块链也只是组合而已
The contract optimization problem falls into the category of difference of convex programing, and is solved by using the iterative convex–concave procedure algorithm.
How to derive the probability distribution of the EV type by SoC estimation techniques.
- Gaussian process regression
I. Introduction
A. Background and Motivation
- smart energy management
- 不协调、可再生能源时断时续
- 为了供需平衡,建设的基础设施代价高昂
- 另一种方法即将电动汽车视为基础设施,让其根据市场信息,决策是否、以何量充放电——IoEV-based DR.
- 两大好处:
- 通信角度:无缝信息收集。
- 能量角度:ev可以作为备用能源库——energy local area networks, virtual power plant
IoEV-based DR面临的挑战:
- 缺少安全的能量交易机制——区块链
- 缺少incentive-compatible DR mechanism: range anxiety——契约论
- information asymmetric: the probability distribution of each type——机器学习GPR估计SoC
- 区块链方面的研究:every transaction is recorded in a verifiable and permanent way.
- 为了避免solve proof-of-work puzzles带来的计算压力,利用consortium blockchain
- SoC estimation
- physical circuit model
- machine learning
- the GPR scheme to obtain the probability distribution of the SoC based on current, voltage, and temperature measurements.
详见08074792(IEEE)
B. Contributions
基本依照challenges
- use the take-it-or-leave contract as performance benchmarks
详见06342942(IEEE)
II. Consortium Blockchain-based Secure Energy Trading
A. System Model
- two entities: EVs and LEAG
- 两者所扮演的角色、调度过程中两者的互动
- 从区块链的角度,看LEAG的三大组分
B. Implementation
- 系统初始化:设定加密算法、ev的注册
- leag设计contract,\((L_k, R_k)\),具体介绍
- 交易与区块链的交互
III. Contracted-based Incentive-Compatible Demand Response for IoEV
A. EV Type Modeling
- 假定是离散且有限的
- Def 1.
- 依据放电能力,确定type的表达式
- type的概率分布的定义
B. Contract Formulation
下面就不梳理结构了,后面读文献每当读完一节就梳理下
- 注意除了IC、IR之外,卖家也有一个约束
- 对于ev,放电时是有额外cost的
- Def 2. IR、IC、monotonicity constraints
- The IC constraint ensures the self-revealing property of the contract
- Lemma 1: reward是type的单调递增函数
- Lemma 2: reward、workload(只是主张)、utility of ev都是type的单调递增函数
证明详见07110552(IEEE)
C. Optimal Contract Design under Information Asymmetry
1) Contract Feasibility
- Theorem 1. Contract Feasibility
2) Problem Transformation
- proof of Theorem 1.
- 这一步主要是减少constraint的数目
3) Optimal Contract with Reduced Constraints
the objective function is concave.
but the IC constraints involve the difference of two concave functions
use the iterative convex–concave procedure algorithm
Variations and extension of the convex–concave procedure
Theorem 2. Convergence
D. Optimal Contract Design without Information Asymmetry
What will happen?
- Proposition 1. the payoff for any EV is zero.
- Proposition 2. reward is fixed regardless of the type of EV
IV. Computational Intelligence-based SoC Estimation
- type决定于三个参数
- 未来行驶距离可以由历史行车轨迹习得
- 电池最大电量是确定的
- SoC要进行估计:it is difficult to directly measure the value of SoC because the energy is stored in a chemical form, and a battery system exhibits highly nonlinear features
- the GPR-based SoC estimation method
- the off-line training stage and the online estimation stage
V. Computational Complexity Analysis
VI. Simulation Results
- the take-it-or-leave contract is a threshold-based contract
- just one contract item, and individual whose type is lower than the threshold type will refuse to sign the contract.
some conclusions(partial):
- 对各种theorem和proposition的验证
- 信息不对称有利于社会福利最大化
- LEAG剥削所得不能补偿ev的损失
- SoC估计误差与优化目标函数值的关系