关于「高意向线索派单:响应衰减、厄朗排队与公平轮转」,《分层公地与储备金精算》给出的定量结论是什么?

结论:第五章用厄朗 C 式排班,三人加到四人,线索剩余热度从 0.8802 升到 0.9728;同一队列、空闲最久者接单,期望成交 384.5 宗,严格轮转只有 353.4 宗。

数据口径:大修支出块状、时点随机、偶有重尾意外;示意大楼即使管理费随通胀上调,按一成计提,三十年内见底机会仍有 0.5715,计提恰好等于平均支出,也有 0.3479。 书中楼盘、线索与参数全为合成或示意;全部计算由书中程序生成;本书是非个人化分析,不构成法律、税务或投资建议,法条与比例以上架前人工复核为准;线索只处理经同意者。

English version

对照数据表

下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库与结论吻合的打印值: 0.8802, 0.9728

A. load A2.0 Erlangs
agents P(wait) mean wait min P(wait<5 min) value left
30.4444 4.44 0.7304 0.8802
40.1739 0.87 0.9360 0.9728
50.0597 0.20 0.9867 0.9934
60.0180 0.05 0.9976 0.9985
70.0048 0.01 0.9996 0.9997
target P(wait<5 min) >= 0.90: smallest m = 4
B. square-root staffing m = A + beta sqrt(A)
beta 0.5: m3, P(wait) 0.4444
beta 1.0: m4, P(wait) 0.1739
beta 1.5: m5, P(wait) 0.0597
beta 2.0: m5, P(wait) 0.0597
C. desk hours: share off-hours 0.25, value at 6 h 0.0158
3 agents in hours: blended value 0.6641
5 agents in hours: blended value 0.7490
exponential off-hours wait, mean 6 h: value 0.0831
查看完整验算程序
# Chapter 5. p52: staffing a lead desk with Erlang C.
# Leads arrive as Poisson, lambda = 12 an hour at peak; a first call and
# its log take 10 minutes on average (exponential), mu = 6 an hour, so
# the offered load is A = 2 Erlangs (illustrative). With m agents:
#   C(m, A) = P(wait) = [A^m / m! * m / (m - A)] /
#             [sum_{k<m} A^k / k! + A^m / m! * m / (m - A)]
#   P(W > t) = C exp(-(m mu - lambda) t),  E[W] = C / (m mu - lambda).
# Value left after the wait uses the two-speed curve of p51 (w = 0.9812,
# fast half-life 18.3 min, slow half-life 24 h), whose Laplace transform
# gives E[q(W)] in closed form. Panel C: a quarter of leads arrive when
# the desk is closed and wait 6 hours on average (illustrative).
import math

LAM, MU = 12.0, 6.0
W, KF, KS = 0.9812, math.log(2.0) / (18.3 / 60.0), math.log(2.0) / 24.0


def erlang_c(m, a):
    if m <= a:
        return 1.0
    top = a ** m / math.factorial(m) * m / (m - a)
    low = math.fsum(a ** k / math.factorial(k) for k in range(m))
    return top / (low + top)


def value_left(m):
    c = erlang_c(m, LAM / MU)
    th = m * MU - LAM
    lt = [(1.0 - c) + c * th / (th + k) for k in (KF, KS)]
    return W * lt[0] + (1.0 - W) * lt[1]


if __name__ == "__main__":
    a = LAM / MU
    print("A. load A = {:.1f} Erlangs".format(a))
    print("   agents  P(wait)  mean wait min  P(wait<=5 min)  value left")
    for m in range(3, 8):
        c = erlang_c(m, a)
        th = m * MU - LAM
        print("   {:>6}  {:.4f}   {:>13.2f}  {:>14.4f}  {:>10.4f}".format(
            m, c, 60.0 * c / th, 1.0 - c * math.exp(-th * 5.0 / 60.0),
            value_left(m)))
    m = 1
    while (1.0 - erlang_c(m, a) * math.exp(-(m * MU - LAM) * 5.0 / 60.0)
           < 0.90):
        m += 1
    print("   target P(wait<=5 min) >= 0.90: smallest m = {}".format(m))
    print("B. square-root staffing m = A + beta sqrt(A)")
    for beta in (0.5, 1.0, 1.5, 2.0):
        m = math.ceil(a + beta * math.sqrt(a))
        print("   beta {:.1f}: m = {}, P(wait) {:.4f}".format(
            beta, m, erlang_c(m, a)))
    off = 0.25
    night = W * math.exp(-KF * 6.0) + (1.0 - W) * math.exp(-KS * 6.0)
    print("C. desk hours: share off-hours {:.2f}, value at 6 h"
          " {:.4f}".format(off, night))
    for m in (3, 5):
        mix = (1.0 - off) * value_left(m) + off * night
        print("   {} agents in hours: blended value {:.4f}".format(m, mix))
    lt = [1.0 / (1.0 + k * 6.0) for k in (KF, KS)]
    exp_night = W * lt[0] + (1.0 - W) * lt[1]
    print("   exponential off-hours wait, mean 6 h: value {:.4f}".format(
        exp_night))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《分层地契公地博弈、储备金破产精算与高意向置业线索派单母典》第 5 章「高意向线索派单:响应衰减、厄朗排队与公平轮转」(版本 1.0.0). 5D Library. DOI 待发布

DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 5DL-B-335 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026335,
  author    = {Fan, Yafu},
  title     = {Strata-Title Commons Games, Sinking-Fund Ruin Actuarial Science and High-Intent Property Lead Routing},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论