● Fleet prediction: demand time series model, courier-order relation Logit Regression model (Java)● Period-based courier scheduling: fleet assignment considering operation needs and courier intents● Result collection: measure the merits of suggested scheduling results by satisfaction of periods● 运力预测：单量预测时间序列模型、人单匹配关系逻辑回归（Java实现）● 分时段排班运筹模型：考虑运营规则与骑手志愿的周维度运力分配（ODPS中使用ORTools）● 结果回收：以时段满足率为核心指标衡量推荐排班、手动排班的优劣，物流完成率整体提升

● Dept. of Intelligent Decision: CBD real-time chain reaction, mid-long term demand supply coordination● Pressure Adjustment: define pressure ratio, realization of flow adjustment and logistics adjustment● Courier scheduling problem: demand estimation, courier portrait and courier scheduling● Customer-side delivery fee: comparison of delivery fee with counterparts, reduce lose rate● Refinement of map-end Information: courier track analysis in electronic fence, DBSCAN clusters● 决策智能部：以CBD实时联动，中长期供需协同为目标的算法开发、工程迭代、结果回收分析● 压力调控：定义商圈压力系数，实现流量测调控(Rank)与物流侧调控(加T/调C配/下补贴/配送圈)● 运力排班：单量预测，骑手画像，人单匹配模型，基于运营规则和骑手志愿的运力排班● 智能C端配送费：与竞对比较C配，通过降C配的方式降低落败率，同时平衡物流总收入● 地图末端精细化：电子围栏内的骑手轨迹分析，通过聚簇算法等得出AOI配送难度指标

● CBD real-time chain reaction: emphasis on logistics adjustment to on flow adjustment, refine the utilization circumstances, and protect high value retailers● Area pressure ratio: (old) the ratio of real-time on-duty load with total load capacity; (new) the prediction of logistics target in the next time slices● Logistics adjustment: T plus strategy, C plus strategy, suspend allowances, limit the delivery range● Flow adjustment: retailer stratification, retailer sequence index, single-retailer stuck adjustment● CBD实时联动：从物流侧调控转变为流量侧调控，精细化调控使用场景，保护高价值商户● 商圈压力系数：(旧)实时背单数量与背单能力的比值；(新)对于将来时间片内的物流核心指标的预测● 物流侧调控：加T，加C配(精细化)，下补贴，缩配送范围；调控策略的制定，考虑不同压力系数档位下的物流核心指标（Java实时线上服务）● 流量侧调控：通过商户分层(V等级)和物流因素(T等级)形成商户排序因子，单店卡餐调控作为辅助

● VRPTW: use AMAP API to get distance and time matrix, solve the problem by Tabu Search, utilize a variety of neighborhood search rules, and consider the 3rd ring restrictions, node-fixed routes, buffered capacity, etc.● VRPTW问题：使用高德地图API获取时间距离矩阵，禁忌搜索算法建模求解，应用多种邻域搜索方式，实际业务中考虑三环限行、主线路、弹性额载等需求（Tabu Search）

● Long-term problem: based on OD and frequency information, optimize the flight plan for a future season; consider the slots, fixed connections, schedule consistency, minimize long connections, etc.● Short-term problem: based on an original flight plan, take the overall contribution as KPI, optimize the daily fleet assignment and aircraft routing, and consider the situation when regulation design defects● 中长期编排：通过航班班次量输入，制定各航司航季网络规划；核心为时空网络模型，再通过建模解决实际业务中的时刻库、固定衔接、班期一致性、减少长过站等要求（使用Gurobi解MIP）● 短期编排：以初始航班计划为基础，日循环进行的机型调整与飞机路线规划（使用CPLEX解MIP）

● Project life-cycle support: POC, algorithm implementation plan, code delivery, solution improvement● Participate in: Airline (FAM), Logistics (VRPTW), Warehousing (location, layout, goods allocation), APS● Fleet Assignment Problem: adopt the time-space network model to solve airline-based FAM problems● Vehicle Routing Problem: multi-vehicle problems with time windows and operation requirements● Inventory Problem: cell-based warehouse design and cargo allocation optimization problems● 全程负责项目POC，算法实施方案，编写交付，后期调优等项目完整生命周期的算法支持工作● 参与项目：航空(FAM)，物流(VRPTW)，仓储(选址问题、布局问题、货位问题)，排产(APS)● 航空网络问题：时空网络建模，解决航网自动编排、时刻时长调整、优化备份运力、机型配置问题● 路径规划问题：使用Tabu Search解决多车型含时间窗VRP，考虑限行、主线路、弹性额载等需求● 库存管理问题：栅格化的仓库货架地堆布局设计，大件货物库内货位优化

● Designed the SEM promotion platform, SEM data platform and frontstage and backstage projects ● 负责前后端产品和SEM营销平台、数据平台：需求分析，前期调研，方案设计，测试上线

● Analyzed the business needs of a SOE publication retailer to build a comprehensive online shopping website;● Wrote bidding documents and demand documents for UI/UE Designs, B2C Frontstage and Statistics Platform;● Worked for Apple + IBM partnership, Telecomm Sector.● 分析某图书出版集团的业务需求，为其建设多元化产品的电商网站● 负责UI/UE设计，B2C前台模块，经营分析模块；撰写招标文件和需求规格说明书● Apple + IBM合作项目部门，电信行业组

● Led weekly discussion sessions in Engineering Graphics & Design, taught Autodesk Revit at the lab;● Led weekly review sessions in Production Planning & Control● 组织＜工程制图学＞的视图概念教学，上机教授制图软件；● 组织＜生产控制与规划＞每周的讨论课

Project: Robust Aircraft Routing Problems in Large Scale Sets, with Prof. Lavanya Marla● Research on aircraft routing under uncertainty, characterize the uncertainty into parameters, perform dataset computations, solve the model and compare the solution robustness● 基于全美航班的延误历史数据，对某航司的机队进行飞机路线规划；建立多套数学模型，将不确定性刻画为参数值，应用多种模型求解与比较最终结果鲁棒性

● Worked as primary assistant of a chemical distributer project and a logistics company project;● Researched overall chemical industry and understood the operational modes and key competences;● Researched the natural gas & shale gas market in China.● 全程参与化工贸易国企战略分析项目，研究化工行业的经营模式与痛点● 国内天然气、页岩气市场研究

● Researched the sales mode of security electronics in a security electronics marketing project;● Estimated the overall demand of distributors and contractors in terms of the different product segments.● 外资500强电子工业产品企业的经销商渠道梳理项目● 对于不同的商品种类，统计估算批发商与经销商的总需求