我用截面数据跑内生转换模型，约3000左右的样本量，movestay的时候，一直迭代不出结果如何解决？
Iteration 88:  log likelihood = -23662.607  (not concave)
Iteration 89:  log likelihood = -23662.539  (not concave)
Iteration 90:  log likelihood = -23644.933  (not concave)
Iteration 91:  log likelihood = -23644.699  (not concave)
Iteration 92:  log likelihood = -23644.501  (not concave)
Iteration 93:  log likelihood = -23644.464  (not concave)
Iteration 94:  log likelihood = -23644.371  (not concave)
Iteration 95:  log likelihood = -23644.358  (not concave)
Iteration 96:  log likelihood = -23644.328  (not concave)
Iteration 97:  log likelihood = -23644.319  (not concave)
Iteration 98:  log likelihood = -23644.293  (not concave)
Iteration 99:  log likelihood =  -23644.21  (not concave)
Iteration 100: log likelihood = -23644.154  (not concave)
Iteration 101: log likelihood = -23642.888  (not concave)
Iteration 102: log likelihood = -23642.638  (not concave)
Iteration 103: log likelihood = -23642.451  (not concave)
Iteration 104: log likelihood = -23640.206  (not concave)
Iteration 105: log likelihood = -23638.874  (not concave)
Iteration 106: log likelihood =  -23638.63  (not concave)
Iteration 107: log likelihood = -23637.727  (not concave)

……