Python Scipy最小化不工作

Question

我试图最小化向量x，y，但它们只是满足约束，没有最小化的工作。 例如： input init:x=[0.2,0.3,0.5]（元素之和为1），feedback:res.x=[0.,0.3,0.5]，它一点都没变！你知道吗

# -*- coding:utf8-*-

import random
import numpy as np
from scipy import optimize
import networkx as nx
import matplotlib.pyplot as plt

def Ud(x, X, Aj, G, p):
    lost = G.node[Aj[-1]]["weight"]
    Aj = Aj[:-1]
    P = 1
    sum = 0

    for xi in x:
        for Xi in X:
            N = set(Aj).intersection(set(Xi))
            # N=[random.randint(90,120) for _ in range(0,1)]
            for n in N:
                P *= (1 - p[n - 1])
        sum += xi * P
        P = 1
    return -lost * sum
    ### objective function for defenders ###
def min_Ud(x, X, A, G, p):
    min = float("inf")
    for Aj in A:
        temp_min = Ud(x, X, Aj, G, p)
        if temp_min < min:
            min = temp_min
    return -min
    ### objective function for attackers ###
def Ua(a, X, A, G, p):
    sum = 0
    P = 1
    for aj, Aj in a, A:
        for Xi in X:
            N = set(Aj[:-1]).intersection(set(Xi))
            for n in N:
                P *= (1 - p[n - 1])
        sum += G.node[Aj[-1]]["weight"] * aj * P
        P = 1
    return sum

### fun for LP ###
def coreLP(X, A, G, p):
    x0 = np.array([0.7, 0.2, 0.1])

    a0 = np.array([0.5, 0.2, 0.3])
    x_res = float("inf")
    a_res = float("inf")

    def c1(x):
        return x.sum() - 1

    cons = ({'type': 'eq', 'fun': c1})

    Ud_star = optimize.minimize(min_Ud, x0, args=(X, A, G, p), constraints=cons, bounds=((0, 1), (0, 1), (0, 1)))
    Ua_star = optimize.minimize(min_Ud, a0, args=(X, A, G, p), constraints=cons, bounds=((0, 1), (0, 1), (0, 1)))
    print Ud_star
    print Ua_star

    return Ud_star.x, Ua_star.x