Unverified50% confidenceFactExact time
标准卡尔曼滤波假设系统状态转移与观测噪声服从高斯分布且系统动力学为线性，在目标加速转弯或被部分遮挡时会导致预测偏差
1
Sources
50%
Confidence
Long-term
Relevance
7/19/2026
First Seen
Sources
Related Claims
Unverified卡尔曼滤波通过建立系统状态空间模型，将预测值与测量值按各自噪声协方差加权融合，是处理IMU漂移问题的经典方案75% similarUnverifiedKalman filtering is a recursive Bayesian estimation algorithm capable of estimating system states in real time from noisy sensor data.74% similarUnverified卡尔曼滤波是线性高斯假设下的贝叶斯最优估计器，其预测-更新两步迭代对应先验与后验概率的递推计算71% similarUnverified卡尔曼滤波核心思想是将预测值与观测值按各自的不确定性（协方差）加权融合，不确定性越低的来源获得越高权重70% similarUnverified过拟合的统计原理是当模型参数量相对训练数据过多时，模型会记住噪声而非学到规律68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/563705API
curl https://kongchang.com/api/v1/knowledge/claims/563705MCP
get_claim(id=563705)