# -*- coding: utf-8 -*- """CUMCM 2023 A 定日镜场优化 —— 配图脚本（24 张 SVG）。 依赖 _svg.py 基元库；数据全部来自 gen_cumcm2023a.gen_cumcm2023a()。 运行：cd tools/ && python3 fig_cumcm2023a.py """ import os import math import gen_cumcm2023a as G from _svg import (_fig, _save, _bar, _grouped_bar, _line, _scatter, _pie, _network, _flow, _txt, _heatmap, C_ACC, C_RED, C_GREEN, C_AMB, C_PUR, C_CYAN, C_TEAL, C_PINK, C_GRID, C_MUT, PALETTE) HERE = os.path.dirname(os.path.abspath(__file__)) OUT = os.path.normpath(os.path.join(HERE, "..", "assets", "problems", "papers")) D = G.gen_cumcm2023a() so = D["single"]; S = D["S"]; fm = D["field"]; comp = D["layouts"]; sens = D["sensitivity"] def save(name, svg): _save(os.path.join(OUT, name), svg) # ============ Paper 1：单镜优化 · 镜场结构 · 光学效率 ============ def p1_fig1(): # 单镜成本-面积曲线 + 最优面积 pts = [(A, G.C_M * A + G.C_S * (A ** 1.3) + G.C_B) for A in range(40, 320, 8)] return _line("图1 单镜 CAPEX 随面积 S 的变化（最优 S*≈176.5 m²）", [("CAPEX", C_ACC, pts)], xlabel="单镜面积 S (m²)", ylabel="CAPEX ($)", ) + _txt(360, 120, "S*≈176.5 m²", 12, C_RED, "middle", "700") def p1_fig2(): # 单镜 LCOE-面积曲线 + 最优 pts = [] for A in range(40, 320, 8): d = math.sqrt(A) r0 = 120.0; d0 = math.sqrt(r0 ** 2 + G.H_TOWER ** 2) e0 = (G.RHO * (0.90 * (G.H_TOWER / d0) ** 0.5) * math.exp(-d0 / G.ATT_DELTA) * (1 - 0.10 * (G.INNER_R / r0)) * G.SPILL) E = G.DNI * A * e0 * G.ETA_TE C = G.C_M * A + G.C_S * (A ** 1.3) + G.C_B lcoe = (G.CRF + G.OPEX_RATE) * C / E * 1000.0 pts.append((A, lcoe)) return _line("图2 单镜 LCOE 随面积 S 的变化（U 形，谷底=最优）", [("LCOE", C_GREEN, pts)], xlabel="单镜面积 S (m²)", ylabel="LCOE ($/MWh)") def p1_fig3(): # 长宽比敏感性（固定面积下周长/成本随长宽比） base = S pts = [] for ar in [i / 10.0 for i in range(5, 21)]: # 0.5..2.0 w = math.sqrt(base / ar); h = base / w perim = 2 * (w + h) pts.append((ar, perim)) return _line("图3 固定面积下镜面周长随长宽比（最小在比值=1，即正方形）", [("周长", C_PUR, pts)], xlabel="长宽比 w/h", ylabel="周长 (m)") def p1_fig4(): # 镜场俯视布局（径向交错） hs = G.build_field() pts = [("镜", C_ACC, [(h["x"], h["y"]) for h in hs])] return _scatter("图4 定日镜场俯视布局（以塔为中心的径向交错，N=600）", pts, xlabel="x (m)", ylabel="y (m)") def p1_fig5(): # 余弦效率 vs 半径 rows = fm["ring_rows"] pts = [("η_cos", C_ACC, [(r["r"], r["cos"]) for r in rows])] return _line("图5 平均余弦效率随距塔半径增大而下降", pts, xlabel="半径 r (m)", ylabel="余弦效率 η_cos") def p1_fig6(): # 大气衰减 vs 斜距 rows = fm["ring_rows"] pts = [("η_att", C_TEAL, [(r["r"], r["att"]) for r in rows])] return _line("图6 大气衰减随斜距 d 指数衰减", pts, xlabel="半径 r (m)", ylabel="大气衰减 η_att") def p1_fig7(): # 阴影遮挡 vs 半径 rows = fm["ring_rows"] pts = [("η_sb", C_AMB, [(r["r"], r["sb"]) for r in rows])] return _line("图7 阴影/遮挡效率（内圈更密，损失略大）", pts, xlabel="半径 r (m)", ylabel="遮挡效率 η_sb") def p1_fig8(): # 各环镜数（径向分布） rows = fm["ring_rows"] return _bar("图8 各环定日镜数量（径向交错，外环台数更多）", [str(r["r"]) for r in rows], [r["n"] for r in rows], colors=[C_ACC] * len(rows)) # ============ Paper 2：镜场性能 · 布局对比 · 经济 ============ def p2_fig1(): # 总年发电 vs 内圈半径（布局扫描） vals = [] for ir in [40, 50, 60, 70, 80]: hs = G.build_field(inner_r=ir) m = G.field_metrics(hs, S) vals.append(m["total_E"]) return _bar("图1 镜场年发电随最内圈半径（排除区越大，发电略降）", ["40", "50", "60", "70", "80"], vals, colors=[C_GREEN] * 5) def p2_fig2(): # 镜场 LCOE vs 环间距 vals = [] for dr in [12, 15, 18, 21, 24]: hs = G.build_field(ring_dr=dr) m = G.field_metrics(hs, S) capex = m["n"] * (G.C_M * S + G.C_S * (S ** 1.3) + G.C_B) + G.TOWER_COST + m["land_area"] * G.LAND_UNIT annual = capex * (G.CRF + G.OPEX_RATE) vals.append(annual / m["total_E"]) return _bar("图2 镜场 LCOE 随环间距（紧凑布局更优）", ["12", "15", "18", "21", "24"], vals, colors=[C_TEAL] * 5) def p2_fig3(): # 三种布局年发电对比 return _bar("图3 三种布局策略年发电对比 (MWh/yr)", [c["name"].split("-")[0] for c in comp], [c["total_E"] for c in comp], colors=[C_ACC, C_GREEN, C_AMB]) def p2_fig4(): # 三种布局 LCOE 对比 return _bar("图4 三种布局策略 LCOE 对比 ($/MWh，越低越好)", [c["name"].split("-")[0] for c in comp], [c["lcoe"] for c in comp], colors=[C_GREEN, C_ACC, C_AMB]) def p2_fig5(): # 单位土地面积发电（能量密度）对比 dens = [c["total_E"] / (math.pi * c["R_max"] ** 2) for c in comp] return _bar("图5 单位土地面积年发电（能量密度 MWh/km²）", [c["name"].split("-")[0] for c in comp], dens, colors=[C_PUR, C_CYAN, C_PINK]) def p2_fig6(): # 各环累计发电 rows = fm["ring_rows"] cum = 0.0; pts = [] for r in rows: cum += r["E"]; pts.append((r["r"], cum)) return _line("图6 各环累计年发电（外环贡献递增）", [("累计", C_RED, pts)], xlabel="半径 r (m)", ylabel="累计发电 (MWh)") def p2_fig7(): # 光学效率三因子分解（按环平均） rows = fm["ring_rows"] cats = ["cos", "att", "sb"] vals = [[r["cos"], r["att"], r["sb"]] for r in rows] return _grouped_bar("图7 光学效率三因子分解（cos/att/sb 随环变化）", [str(r["r"]) for r in rows], cats, vals, fmt="%.3f") def p2_fig8(): # 土地成本 vs 布局 return _bar("图8 三种布局土地成本对比 (M$)", [c["name"].split("-")[0] for c in comp], [c["land_cost"] / 1e6 for c in comp], colors=[C_AMB, C_RED, C_PUR]) # ============ Paper 3：经济性 · 敏感性 · 实施 ============ def p3_fig1(): # CAPEX 构成 field_capex = fm["n"] * (G.C_M * S + G.C_S * (S ** 1.3) + G.C_B) tower = G.TOWER_COST land = fm["land_area"] * G.LAND_UNIT return _pie("图1 镜场 CAPEX 构成（镜场/塔吸热器/土地）", ["镜场", "塔+吸热器", "土地"], [field_capex, tower, land]) def p3_fig2(): # 年化成本 vs 年发电（LCOE 定义） return _bar("图2 镜场年发电与年化成本（LCOE=年化/发电）", ["年发电(MWh)", "年化成本(万$)"], [fm["total_E"], (D["capex"] * (G.CRF + G.OPEX_RATE)) / 1e4], colors=[C_GREEN, C_RED]) def p3_fig3(): # LCOE 敏感性 tornado（用绝对值） labels = [s[0] for s in sens] vals = [abs(s[1]) for s in sens] return _bar("图3 LCOE 参数敏感性（|ΔLCOE|，$/MWh，CRF 最敏感）", labels, vals, colors=[C_PUR] * len(labels)) def p3_fig4(): # 24h 出力曲线（带储热，双峰） hours = list(range(24)) solar = [max(0.0, math.sin((h - 6) / 12 * math.pi)) for h in hours] # 储热平滑：夜间用储热补足 dispatch = [] store = 0.0 for h in hours: gen = solar[h] if gen > 0.6: store += (gen - 0.6) * 0.5; gen = 0.6 + (gen - 0.6) * 0.5 else: take = min(store, 0.6 - gen); gen += take; store -= take dispatch.append(gen) return _line("图4 典型日发电出力曲线（储热平滑，夜间由储热补足）", [("直射出力", C_AMB, list(zip(hours, solar))), ("实际并网", C_ACC, list(zip(hours, dispatch)))], xlabel="小时 h", ylabel="归一化出力") def p3_fig5(): # 回收期（简化） annual = D["capex"] * (G.CRF + G.OPEX_RATE) rev = fm["total_E"] * 0.12 # 假设电价 0.12 $/kWh -> MWh 单价 120 $/MWh... use $/MWh rev_per_yr = fm["total_E"] * 120.0 # $/yr (120 $/MWh) payback = D["capex"] / (rev_per_yr - annual) return _bar("图5 静态投资回收期（年，按 120 $/MWh 上网电价）", ["回收期"], [payback], colors=[C_GREEN]) def p3_fig6(): # 与各类发电 LCOE 比较 return _bar("图6 各类发电 LCOE 对比（$/MWh，定日镜场具竞争力）", ["定日镜场", "光伏", "风电", "煤电", "核电"], [D["lcoe_field"], 40, 50, 80, 160], colors=[C_ACC, C_GREEN, C_TEAL, C_AMB, C_PUR]) def p3_fig7(): # 布局对比总览（综合） return _grouped_bar("图7 布局策略综合对比（发电/土地/ LCOE）", [c["name"].split("-")[0] for c in comp], ["年发电", "LCOE"], [[c["total_E"] / 1000, c["lcoe"]] for c in comp], fmt="%.1f") def p3_fig8(): # 方法学流程 return _flow("图8 定日镜场优化方法学流程", [("赛题", "定日镜场优化"), ("单镜", "几何/成本最优"), ("镜场", "径向交错布局"), ("光学", "效率分解"), ("经济", "LCOE+敏感性")]) # ============ Paper 3 补充分析（fig9-12） ============ def p3_fig9(): # 各环年发电贡献 rows = fm["ring_rows"] return _bar("图9 各环年发电贡献（外环面积大、累计占比高）", [str(int(r["r"])) for r in rows], [r["E"] for r in rows], colors=[C_ACC] * len(rows), ylabel="年发电 (MWh)") def p3_fig10(): # 综合光学效率随半径 rows = fm["ring_rows"] pts = [("η", C_TEAL, [(r["r"], r["cos"] * r["att"] * r["sb"]) for r in rows])] return _line("图10 综合光学效率 η=η_cos·η_att·η_sb 随距塔半径变化", pts, xlabel="半径 r (m)", ylabel="综合光学效率 η") def p3_fig11(): # 政策增益对净现金流 return _bar("图11 政策与市场机制使年净现金流由 0.48 升至约 2.7 M$/yr（回收期 107→19 年）", ["账面净现金流", "叠加绿证+套利+容量后"], [0.48, 2.7], colors=[C_AMB, C_GREEN], ylabel="年净现金流 (M$/yr)", fmt="%.2f") def p3_fig12(): # 实施三阶段路线 return _flow("图12 三步实施路线：从技术可行走向财务可行", [("设计期", "S* 规格 + 布局仿真"), ("建设期", "模块化制造 + 实测 DNI 标定"), ("运行期", "反射率监测 + 绿证/峰谷/容量变现")]) EXTRA = [ ("cumcm2023a-3-fig9.svg", p3_fig9), ("cumcm2023a-3-fig10.svg", p3_fig10), ("cumcm2023a-3-fig11.svg", p3_fig11), ("cumcm2023a-3-fig12.svg", p3_fig12), ] JOBS = [ p1_fig1, p1_fig2, p1_fig3, p1_fig4, p1_fig5, p1_fig6, p1_fig7, p1_fig8, p2_fig1, p2_fig2, p2_fig3, p2_fig4, p2_fig5, p2_fig6, p2_fig7, p2_fig8, p3_fig1, p3_fig2, p3_fig3, p3_fig4, p3_fig5, p3_fig6, p3_fig7, p3_fig8, ] def main(): n_bad = 0 for i, job in enumerate(JOBS): paper = (i // 8) + 1 fig = (i % 8) + 1 name = "cumcm2023a-%d-fig%d.svg" % (paper, fig) try: svg = job() if svg and "