供应链优化器
用于供应链分析、库存管理、物流优化和供应商评估的综合框架。
供应链映射模板
端到端供应链图
RAW MATERIALS → SUPPLIERS → MANUFACTURING → DISTRIBUTION → CUSTOMER
MAPPING STEPS:
1. Identify all nodes (suppliers, plants, warehouses, customers)
2. Map material flows between nodes
3. Map information flows (orders, forecasts, POs)
4. Map financial flows (payments, invoicing)
5. Record lead times at each stage
6. Identify bottlenecks and single points of failure
NODE DETAIL TEMPLATE:
| Node | Type | Location | Lead Time | Capacity | Utilization |
| -------------- | ---------- | --------- | --------- | -------- | ----------- |
| | Supplier | | | | |
| | Plant | | | | |
| | Warehouse | | | | |
| | DC | | | | |
库存优化
经济订货量（EOQ）
EOQ = sqrt(2DS / H)
Where: D = Annual demand, S = Ordering cost/order, H = Holding cost/unit/year
Total Cost = (D/Q)S + (Q/2)H + DC
Example: D=10,000, S=$50, H=$5 → EOQ = 447 units, 22.4 orders/year
安全库存与再订货点
SAFETY STOCK: SS = z x sigma_dLT
REORDER POINT: ROP = (Avg Daily Demand x Lead Time) + Safety Stock
SERVICE LEVEL FACTORS:
| Service Level | z-Score | Use Case |
| ------------- | ------- | ----------------- |
| 90.0% | 1.28 | Basic coverage |
| 95.0% | 1.65 | Standard |
| 99.0% | 2.33 | High service |
| 99.9% | 3.09 | Critical items |
库存 KPI 仪表板
| 指标 | 公式 | 目标 |
|---|
| 库存周转率 | COGS / 平均库存 | 因行业而异 |
| 供货天数 | 平均库存 / (COGS / 365) | 尽可能低 |
| 订单满足率 | 完整履行的订单数 / 总订单数 | 97%+ |
| 缺货率 | 缺货事件数 / 总需求事件数 | < 2% |
| 库存持有成本 % | 持有成本 / 平均库存价值 | 15-30% |
| 滞销库存 % | 无流动商品数 / SKU 总数 | < 5% |
| 库存准确率 | 盘点正确次数 / 盘点总次数 | 99%+ |
| GMROI | 毛利 / 平均库存成本 | > 2.0 |
ABC-XYZ 分析框架
ABC CLASSIFICATION (Value):
A Items: Top 20% of SKUs = ~80% of annual consumption value
→ Tight control, frequent review, accurate forecasts
B Items: Next 30% of SKUs = ~15% of value
→ Moderate control, periodic review
C Items: Bottom 50% of SKUs = ~5% of value
→ Minimal control, simple replenishment rules
XYZ CLASSIFICATION (Demand Variability):
X: Coefficient of Variation < 0.5 → Stable, predictable demand
Y: CV between 0.5 and 1.0 → Some variation, trend/seasonal
Z: CV > 1.0 → Highly irregular, sporadic demand
COMBINED MATRIX:
| Class | AX | AY | AZ |
| ----- | ---------- | ------------ | ------------- |
| Strat | JIT/Kanban | Forecast | Order on demand|
| Class | BX | BY | BZ |
| Strat | Reorder pt | Buffer stock | Min/Max |
| Class | CX | CY | CZ |
| Strat | Bulk buy | Periodic rev | Eliminate? |
供应商评分卡
供应商评估矩阵
| 标准 | 权重 | 评分（1-5） | 加权得分 | 备注 |
|---|
| 质量（PPM） | 25% | | | |
| 交付（OTIF） | 20% | | | |
| 定价 | 20% | | | |
| 响应能力 | 10% | | | |
| 财务状况 | 10% | | | |
| 创新 | 5% | | | |
| 可持续性 | 5% | | | |
| 风险状况 | 5% | | | |
| 总计 | 100% | | __ / 5.0 | |
RATING SCALE:
4.5-5.0 Strategic Partner — expand relationship
3.5-4.4 Preferred Supplier — maintain, develop
2.5-3.4 Approved Supplier — improvement plan required
< 2.5 Probation / Exit — find alternative
供应商绩效跟踪
| KPI | 目标 | Q1 实际值 | Q2 实际值 | Q3 实际值 | Q4 实际值 | 趋势 |
|---|
| 准时交付 | 98%+ | | | | | |
| 质量（PPM） | < 500 | | | | | |
| 交货周期（天） | | | | | | |
| 价格偏差 | +/- 2% | | | | | |
| 响应时间（小时） | < 24 | | | | | |
| 纠正措施 | < 2/qtr | | | | | |
总拥有成本（TCO）
TCO = Acquisition Costs + Operating Costs + Disposal Costs
ACQUISITION COSTS:
Purchase price
+ Shipping / freight
+ Customs / duties / tariffs
+ Procurement labor
+ Quality inspection
+ Supplier qualification
= Total Acquisition
OPERATING COSTS (over useful life):
Maintenance & repair
+ Inventory carrying cost
+ Warranty claims
+ Downtime cost (if component fails)
+ Training / support
+ Quality failures (scrap, rework)
= Total Operating
DISPOSAL COSTS:
Decommissioning
+ Recycling / disposal fees
+ Environmental compliance
= Total Disposal
TCO = Total Acquisition + Total Operating + Total Disposal
TCO 比较模板
| 成本要素 | 供应商 A | 供应商 B | 供应商 C |
|---|
| 单价 | | | |
| 运费 | | | |
| 关税 / 进口税 | | | |
| 质量成本（估算） | | | |
| 库存持有成本 | | | |
| 交货周期成本 | | | |
| 风险溢价 | | | |
| 单位 TCO | | | |
| 年度 TCO | | | |
需求预测方法
| 方法 | 适用场景 | 预测周期 | 所需数据 |
|---|
| 移动平均 | 需求稳定 | 短期 | 3-12 个周期的历史数据 |
| 指数平滑 | 趋势检测 | 短期 | 加权近期数据 |
| Holt-Winters | 季节性模式 | 中期 | 2 年以上的季节性数据 |
| 线性回归 | 带有因果因素的趋势 | 中期 | 需求数据 + 驱动因素 |
| ARIMA | 复杂时间序列 | 短至中期 | 50 个以上的数据点 |
| 机器学习 | 多变量模式 | 任意 | 大型数据集 |
| Delphi / 专家法 | 新产品、突发事件 | 长期 | 专家小组 |
预测准确度指标
MAD (Mean Absolute Deviation):
MAD = (1/n) x SUM(|Actual - Forecast|)
MAPE (Mean Absolute Percentage Error):
MAPE = (1/n) x SUM(|Actual - Forecast| / Actual) x 100
BIAS (Tracking Signal):
Bias = SUM(Actual - Forecast) / MAD
Target: Between -4 and +4
ACCURACY BENCHMARKS:
| Forecast Horizon | Good MAPE | Acceptable MAPE |
| ---------------- | --------- | --------------- |
| 1 month | < 15% | < 25% |
| 3 months | < 25% | < 35% |
| 6 months | < 30% | < 45% |
| 12 months | < 35% | < 50% |
物流优化
运输方式选择
| 方式 | 单位成本 | 速度 | 运力 | 适用场景 |
|---|
| 卡车 | 中 | 快 | 中 | 区域运输、门到门运输 |
| 铁路 | 低 | 慢 | 很高 | 散装货物、国内长途运输 |
| 海运 | 很低 | 很慢 | 很高 | 国际运输、散装货物 |
| 空运 | 很高 | 很快 | 低 | 高价值、紧急、易腐货物 |
| 多式联运 | 低至中 | 中 | 高 | 长途运输、经济高效 |
仓库布局优化
SLOTTING STRATEGY:
Fast movers (A items) → Near shipping dock, prime pick locations
Medium movers (B items) → Middle zones
Slow movers (C items) → High racks, far locations
LAYOUT PRINCIPLES:
1. Minimize travel distance for highest-velocity items
2. Group items frequently ordered together
3. Separate receiving and shipping areas
4. Reserve staging areas for cross-docking
5. Maintain clear aisle widths for equipment
前置时间分析
TOTAL LEAD TIME:
Order Processing Time
+ Supplier Manufacturing Time
+ Transportation Time
+ Receiving / Inspection Time
+ Internal Processing Time
= Total Lead Time
LEAD TIME VARIABILITY:
Track actual vs. quoted lead times over 12+ orders
Calculate standard deviation
Use for safety stock calculations
LEAD TIME REDUCTION STRATEGIES:
| Strategy | Typical Reduction | Effort |
| --------------------- | ----------------- | --------- |
| Vendor-managed inv. | 30-50% | Medium |
| Local sourcing | 40-70% | High |
| Process automation | 10-30% | Medium |
| Blanket POs | 20-40% | Low |
| Consignment stock | 50-80% | Medium |
| 3PL consolidation | 10-25% | Low |
供应链风险评估
风险识别矩阵
| 风险类别 | 风险事件 | 可能性 | 影响 | 评分 | 缓解措施 |
|---|
| 供应 | 单一来源故障 | | | | 双重来源 |
| 需求 | 需求激增／骤降 | | | | 缓冲库存、灵活调整 |
| 物流 | 港口拥堵 | | | | 替代路线 |
| 地缘政治 | 关税、制裁 | | | | 近岸外包 |
| 自然灾害 | 地震、洪水 | | | | 地理多元化 |
| 网络安全 | 系统遭入侵 | | | | 安全协议 |
| 质量 | 批次故障 | | | | 检验、冗余 |
| 财务 | 供应商破产 | | | | 财务监控 |
RISK SCORING:
Likelihood: 1 (Rare) to 5 (Almost Certain)
Impact: 1 (Negligible) to 5 (Catastrophic)
Risk Score = Likelihood x Impact
ACTION THRESHOLDS:
20-25 Critical — immediate action, executive attention
12-19 High — mitigation plan required within 30 days
6-11 Medium — monitor quarterly, contingency plans
1-5 Low — accept or monitor annually
KPI 仪表板模板
| 类别 | KPI | 目标 |
|---|
| 成本 | 供应链成本占收入的比例 | 5-10% |
| 成本 | 每笔订单成本 | 尽可能降低 |
| 服务 | 完美订单率 | 95%+ |
| 服务 | 按时足量交付率（OTIF） | 98%+ |
| 服务 | 客户订单满足率 | 97%+ |
| 效率 | 库存周转率 | 因行业而异 |
| 效率 | 现金周转周期 | 尽可能缩短 |
| 效率 | 仓库利用率 | 85% |
| 效率 | 预测准确度（MAPE） | < 20% |
另请参阅
