← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 9175.c · 生成模型:GPT-6

多个行为叠加不保证解决冲突

理解等级:中级

趋光与避障可按权重组合,便于用少量参数形成运动,但两种作用也可能相互抵消或形成循环。训练或进化能选择权重,却不消除任务和环境覆盖要求;需要通过具体轨迹检查组合行为。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      Behaviors such as phototaxis and obstacle avoidance can also be combined by simply weighing each input accordingly. This idea has been popularized by the neuroscientist Valentino Braitenberg, who augmented this system with additional ideas around basic forms of learning (by changing the weights based on events such as collisions), natural selection (building robots with random weights and selecting those that perform best), and analogies to

    参考来源

    - 位置:《Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)》第11章《任务执行》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)/_epub-src/OEBPS/xhtml/chapter_11.xhtml) - 结论依据:摘录中关于多个行为叠加不保证解决冲突的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。 - 原始内容:Behaviors such as phototaxis and obstacle avoidance can also be combined by simply weighing each input accordingly. This idea has been popularized by the neuroscientist Valentino Braitenberg, who augmented this system with additional ideas around basic forms of learning (by changing the weights based on events such as collisions), natural selection (building robots with random weights and selecting those that perform best), and analogies to