Python /C++ · 信息学奥赛、编程私教，本人是高校教师 有中外教学经验

【思维线】 ▍C1 Python 创意编程入门（12 次课，零基础） 从第一行 print 写到第一个自己设计的小游戏。内容覆盖变量、分支、循环、函数、列表与字典、 文件读写，每节课都有可运行的小产出，不讲空概念。 学完能做什么：独立写出 300 行以内的 Python 程序，完成 3 个小项目（猜数字 / 文字冒险 / 成绩统计小工具），不再害怕报错信息。 ▍C2 Python 项目实战（16 次课） 从"会写"到"能交付"。学习代码组织与调试方法、常用第三方库（数据处理、绘图、网络请求）， 理解什么叫"写得好的代码"。 学完能做什么：独立完成 2 个完整项目（如数据可视化分析工具、带界面的小游戏）， 能讲清楚自己每一行代码在干什么。 ▍C3 从 Python 到 C++：程序设计的第二层（16 次课） 为什么有了 Python 还要学 C++？这一门回答这个问题。讲清楚类型、内存、指针、编译与运行， 用 C++ 重写此前做过的 Python 项目，体会"同一件事的两种写法"。 学完能做什么：能用 C++ 独立实现中等规模程序，理解程序是如何在机器上跑起来的， 为竞赛和后续工科学习打底。 ▍C4 算法与数据结构（24 次课） 编程真正分高下的地方。覆盖线性表、栈与队列、递归与分治、排序与查找、贪心、动态规划、 图论基础，配合在线判题系统（OJ）实战。 学完能做什么：掌握 20+ 经典算法模板，能在 OJ 上独立解普及组难度题目， 为蓝桥杯和信息学奥赛铺路。 ▍C5 让电脑"看见"：机器视觉与 AI 入门（12 次课） 用现成的模型做真实的事。讲图像处理基础、目标检测与识别（YOLO 一类）， 带孩子跑通"输入一张图 → 输出结果"的完整链路。 学完能做什么：完成 2 个人工智能小项目（如人脸检测、物品识别、答题卡识别）， 理解 AI 能做什么、不能做什么，避免停留在"AI 很神奇"的层面。 ▍C6 让代码动起来：硬件与机器人编程（16 次课） 代码不止在屏幕里。使用单片机（STM32 / Arduino 类）搭配传感器、电机与机械结构， 从点亮一盏灯做到能自主避障的小车或机械臂。 学完能做什么：完成 1 个能演示的软硬件综合作品，理解软件与硬件如何协同， 为机器人与嵌入式类竞赛做准备。 【竞赛线】 ▍K1 蓝桥杯 Python / C++ 组集训（20 次课） 按赛程倒推排课：语法查漏 → 真题分类突破（模拟、枚举、搜索、动态规划）→ 限时模拟赛 → 错题复盘。全程用真题和计时训练，不搞泛泛而讲。 目标：省赛奖项，冲刺国赛。 ▍K2 信息学奥赛 CSP-J 冲刺（32 次课） 面向入门组（CSP-J）的完整备考：语法精讲 + 算法体系 + 每周模拟赛 + 代码风格与调试规范。 强调"会做"更要"做对"，减少无谓失分。 目标：入门组一等奖 / 二等奖。 ▍K3 科创项目工坊（挑战杯 / 创新大赛，16 次课） 竞赛不只考写代码，还考"做一个有价值的东西"。从选题、技术方案、原型开发到路演答辩， 全程带做，产出一份能报上去的完整作品。 目标：产出 1 个可参赛的科创作品（含技术报告与答辩材料）。 ▍K4 嵌入式与机器人竞赛集训（24 次课） 面向全国大学生嵌入式芯片与系统设计竞赛、机器人科技创新类赛事的青少年组与预备赛道。 内容为硬件选型、控制程序、传感融合与视觉模块的整合训练。 目标：产出 1 个完整参赛作品并完赛冲奖。 【教学说明】 · 每次课后一份书面反馈：学了什么、卡在哪、下次怎么补；每月一次阶段总结。 · 每 4–6 次课交付一个可演示的项目，进度看得见。 · 讲"为什么"不背"怎么写"，换题型也能自己做出来。 · 课时为建议值，第一节课测评后按孩子实际水平调整。 TRACK 1 — COMPUTATIONAL THINKING ▍C1 Creative Coding with Python (12 sessions · no prerequisites) From the first print statement to a game of your own design. Variables, conditionals, loops, functions, lists and dictionaries, file I/O — every session ends with something that runs, never just theory. By the end: write Python programmes of up to 300 lines independently and complete three mini-projects (a guessing game, a text adventure, a grade-statistics tool) — and no longer fear error messages. ▍C2 Python Project Lab (16 sessions) From "can write" to "can deliver". Code organisation and debugging habits, widely used third-party libraries (data handling, plotting, HTTP), and what "well-written code" actually means. By the end: deliver two complete projects (e.g. a data visualisation tool, a small game with a graphical interface) and explain what every line does. ▍C3 From Python to C++: The Next Layer (16 sessions) Why learn C++ when you already know Python? This course answers that question. Types, memory, pointers, compiling versus running — students rebuild their earlier Python projects in C++ and see the same problem solved two ways. By the end: implement medium-sized programmes in C++ independently and understand what actually happens when a programme runs — groundwork for competitions and for any engineering study later. ▍C4 Algorithms & Data Structures (24 sessions) Where programmers are really separated. Linear structures, stacks and queues, recursion and divide-and-conquer, sorting and searching, greedy methods, dynamic programming, introductory graph theory — all practised on an online judge. By the end: 20+ classic algorithm templates, able to solve entry-level contest problems independently, ready for the Lanqiao Cup and the Olympiad in Informatics. ▍C5 Teaching Computers to See: Machine Vision & AI (12 sessions) Doing real things with ready-made models. Image-processing fundamentals, object detection and recognition (YOLO-class models), running the full pipeline from input image to usable output. By the end: two AI mini-projects (face detection, object recognition, answer-sheet scanning) and a grounded sense of what AI can and cannot do — instead of vague awe. ▍C6 Making Code Move: Hardware & Robotics Programming (16 sessions) Code that does not stay on the screen. Microcontrollers (STM32 / Arduino class) with sensors, motors and mechanical parts — from blinking an LED to a self-navigating obstacle-avoiding car or a robotic arm. By the end: one working hardware-software project to demonstrate, a real feel for how software and hardware cooperate, and preparation for robotics and embedded contests. TRACK 2 — COMPETITION ▍K1 Lanqiao Cup Bootcamp, Python / C++ (20 sessions) Scheduled backwards from the contest date: close grammar gaps → break down past papers by category (simulation, enumeration, search, dynamic programming) → timed mock contests → structured error review. Real problems, real timers, no generic lecturing. Goal: provincial-level award, with a run at the national final. ▍K2 Olympiad in Informatics — CSP-J Sprint (32 sessions) Full preparation for the entry-level division: focused grammar, the full algorithm syllabus, weekly mock contests, and code style with debugging discipline. Solving a problem is not enough — solving it correctly under time pressure is. Goal: first or second prize in the entry-level division. ▍K3 Innovation Project Studio (16 sessions) Contests reward more than code; they reward something worth building. From topic selection and technical design through prototype to pitch and defence — guided end to end, producing a submission-ready piece of work. Goal: one competition-ready innovation project, including technical report and defence materials. ▍K4 Embedded & Robotics Contest Training (24 sessions) For youth divisions and feeder tracks of the National Undergraduate Embedded Chip & System Design Contest and robotics innovation competitions. Hardware selection, control programming, sensor fusion and vision-module integration. Goal: one complete competition entry, finished and submitted. HOW I TEACH · Written feedback after every session: what was covered, where the student got stuck, what comes next. · A demo-able project every 4–6 sessions — progress the family can actually see. · Explain "why", not rote "how", so skills transfer to unfamiliar problem types. · Session counts are indicative; adjusted after a diagnostic assessment in the first lesson.