-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathsection_3-7.tex
executable file
·714 lines (645 loc) · 18.8 KB
/
section_3-7.tex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
\subsection{分段线形插值}
\frame{
$n$次插值是根据给定的$y = f(x)$函数表构造代数多项式$P_n(x)$
\begin{center}
$\Downarrow$
\end{center}
$P_n(x)$为$n$次插值函数,可以近似替代$y = f(x)$。
对于$n$次插值,随着插值节点的增加,插值多项式的次数也会增加。其误差:
\begin{equation*}
E_N (x) = \frac{(x - x_0)(x - x_1) \cdots (x - x_N ) f^{(N+1)}(c)}{(N + 1)!}
\end{equation*}
for some value $c = c(x)$ that lies in the interval $[a, b]$.
\begin{center}
$\Downarrow$
\end{center}
\begin{block}{问题}
插值多项式的次数越高,插值多项式$P_n(x)$与$f(x)$的误差就会越小?
\end{block}
}
\frame{
\frametitle{例}
给定函数
\begin{equation*}
f(x) = \frac{1}{1+25x^2} \ \ \ \ (-1 \le x \le 1)
\end{equation*}
取等间距插值节点$x_i = -1 + \frac{1}{5}i (i = 0, 1, \ldots, 10)$,试建立插值多项式$P_{10}(x)$,并考察$P_{10}(x)$与$f(x)$的误差。
\begin{center}
$\Downarrow$
\end{center}
由Lagrange型$n$次插值公式可得:
\begin{equation*}
P_{10}(x) = \sum_{i=0}^{10} f(x_i) L_i(x)
\end{equation*}
}
\frame{
其中
\begin{equation*}
x_i = -1 + \frac{1}{5}i \ \ \ \ (i = 0, 1, \ldots, n)
\end{equation*}
\begin{equation*}
f(x_i) = \frac{1}{1 + 25x_i^2} \ \ \ \ (i = 0, 1, \ldots, n)
\end{equation*}
\begin{equation*}
L_i(x) = \frac{(x- x_0)\cdots(x- x_{i-1})(x- x_i)\cdots(x- x_10)}{(x_i-x_0)\cdots(x_i-x_{i-1})(x_i-x_{i+1})\cdots(x_i-x_10)} \ \ \ \ (i = 0, 1, \ldots, n)
\end{equation*}
}
\frame{
\frametitle{Runge(龙格)现象}
下图为函数$f(x)= \frac{1}{1+25x^2}$和$y=P_{10}(x)$的图形。
\begin{figure}
\begin{center}
\includegraphics[width=40mm]{chap-3/fig_4-17.png}
\end{center}
\end{figure}
从该图可以观察到:
\begin{itemize}
\item 在原点附近能较好的逼近$f(x)$,
\item 在其他地方$f(x)$和$P_{10}(x)$的误差较大,
\item 越靠近两个端点$-1$和$1$,误差越大,还出现了激烈震荡现象
\end{itemize}
\begin{center}
$\Downarrow$ \\
Runge(龙格)现象
\end{center}
}
\frame{
\begin{center}
Runge(龙格)现象 \\
$\Downarrow$
\end{center}
高次插值多项式并不一定能很好地近似代替被插值函数。
如果给定的插值节点较多,
\begin{center}
$\Downarrow$
\end{center}
\begin{block}{分段插值}
需要把插值区间分为若干段,在每个分段上使用低次插值多项式来近似代替函数$f(x)$
\end{block}
}
\frame{
\frametitle{分段线性插值}
根据给定的函数$f(x)$的函数表
\begin{table}
\begin{tabular}{| c | c | c | c |}
\hline
$x$ & $x_0$ & $\cdots$ & $x_n$ \\
\hline
$y = f(x)$ & $y_0$ & $\cdots$ & $y_n$ \\
\hline
\end{tabular}
\end{table}
构造函数$P(x)$满足条件:
\begin{itemize}
\item $P(x)$在$[x_i, x_{i+1}]$ $(i = 0, 1, \ldots, n-1)$上为不超过一次的代数多项式
\item $P(x_i) = y_i$ ($i = 0, 1, \ldots, n$)
\end{itemize}
\begin{center}
$\Downarrow$
\end{center}
$P(x)$为分段线性插值函数。
}
\frame{
\frametitle{分段线性插值函数}
分段线性插值的目的:分段构造$P(x)$来近似代替$f(x)$。
\begin{equation*}
P(x) = \left\{
\begin{array}{c c}
\frac{x -x_1}{x_0 - x_1} y_0 + \frac{x -x_0}{x_1 - x_0} y_1 & (x_0 \le x \le x_1) \\
\frac{x -x_2}{x_1 - x_2} y_1 + \frac{x -x_1}{x_2 - x_1} y_2 & (x_1 \le x \le x_2) \\
\vdots & \vdots \\
\frac{x -x_n}{x_{n-1} - x_n} y_{n-1} + \frac{x - x_{n-1}}{x_n - x_{n-1}} y_n & (x_{n-1} \le x \le x_n) \\
\end{array}
\right.
\end{equation*}
}
\frame{
\frametitle{分段线性插值函数也可以由插值基函数组合而成}
设$L_0(x)$, $L_1(x)$, $\cdots$, $\L_n(x)$为线性函数:
\begin{equation*}
L_0(x) = \left\{
\begin{array}{l l}
\frac{x -x_1}{x_0 - x_1} & (x_0 \le x \le x_1) \\
0 & (x_1 < x \le x_n) \\
\end{array}
\right.
\end{equation*}
\begin{equation*}
L_i(x) = \left\{
\begin{array}{l l}
\frac{x -x_{i-1}}{x_i - x_{i-1}} & (x_{i-1} \le x \le x_i) \\
\frac{x -x_{i+1}}{x_i - x_{i+1}} & (x_i < x \le x_{i+1}) \\
0 & x \in [a, b] - [ x_{i-1}, x_{i+1} ] \ \ \ ( i = 1, 2, \ldots, n-1 ) \\
\end{array}
\right.
\end{equation*}
\begin{center}
$\vdots$
\end{center}
\begin{equation*}
L_n(x) = \left\{
\begin{array}{l l}
\frac{x -x_{n-1}}{x_n - x_{n-1}} & (x_{n-1} \le x \le x_n) \\
0 & (x_0 < x < x_{n-1}) \\
\end{array}
\right.
\end{equation*}
}
\frame{
分段线性插值函数的另外一种形式:
\begin{equation*}
P(x) = \sum_{i=0}^n y_i L_i(x)
\end{equation*}
\begin{block}{分段线性插值算法}
\begin{itemize}
\item 读入插值节点$x_i$,$y_i$ ($i=0, 1, \ldots, n$)和点$xx$
\item 对$i=0, 1, \ldots, n$做:
\begin{itemize}
\item 如果$xx<x_i$,则
\item $yy = \frac{xx - x_i}{x_{i-1} -x_i} y_{i-1} + \frac{xx - x_{i-1}}{x_i - x_{i-1}} y_i$
\item 输出点$xx$相对应的函数近似值$yy$
\end{itemize}
\item 结束
\end{itemize}
\end{block}
}
\frame{
\begin{block}{定理}
给定如下表所示$y = f(x)$的函数表,令$a = x_0$,$b = x_n$,$f(x) \in C^1 [a, b]$ ,
$f^{(2)}(x)$在$ [a, b]$上存在,
\begin{table}
\begin{tabular}{| c | c | c | c |}
\hline
$x$ & $x_0$ & $\cdots$ & $x_n$ \\
\hline
$y = f(x)$ & $y_0$ & $\cdots$ & $y_n$ \\
\hline
\end{tabular}
\end{table}
$P(x)$是$f(x)$的分段线性插值函数,则有:
\begin{equation*}
|R(x)| = |f(x) - P(x)| \le \frac{h^2}{8} M
\end{equation*}
其中$h = \max_{0 \le i \le n-1} |x_{i+1} - x_i|$;$M = \max_{a \le x \le b} |f^{(2)}(x)|$。
\end{block}
}
\subsection{Hermite插值}
\frame{
在某些实际问题中,为了保证插值函数更好地逼近被插值函数$f(x)$,
\begin{itemize}
\item 插值函数在插值节点上的值与被插值函数$f(x)$在插值节点上的值相等
\item 插值函数在插值节点上的{\Large 导数值}与被插值函数$f(x)$在插值节点上的{\Large 导数值}相等
\end{itemize}
\begin{center}
$\Downarrow$ \\ \vspace{5mm}
{\huge Hermite插值}
\end{center}
}
\frame{
\frametitle{三次Hermite插值}
给定如下表所示$y = f(x)$的函数表,
\begin{table}
\begin{tabular}{| c | c | c |}
\hline
$ x $ & $x_0$ & $x_1$ \\
\hline
$y = f(x)$ & $y_0$ & $y_1$ \\
\hline
$y' = f'(x)$ & $m_0$ & $m_1$ \\
\hline
\end{tabular}
\end{table}
构造函数$H(x)$满足条件:
\begin{itemize}
\item $H(x)$为不超过三次的代数多项式
\item $H(x_0) = y_0$,$H(x_1) = y_1$;$H'(x_0) = m_0$,$H'(x_1) = m_1$
\end{itemize}
\begin{center}
$\Downarrow$ \\ \vspace{5mm}
$H(x)$为三次Hermite插值函数。
\end{center}
}
\frame{
设
\begin{equation*}
H(x) = y_0 h_0(x) + y_1 h_1(x) + m_0 H_0(x) + y_1 H_1(x)
\end{equation*}
其中$h_0(x)$,$h_1(x)$,$H_0(x)$和$H_1(x)$为不超过三次的代数多项式,且满足:
\begin{table}
\begin{tabular}{| c | c c | c c|}
\hline
& 函数值 & & 导数值 & \\
& $x_0$ & $x_1$ & $x_0$ & $x_1$ \\
\hline
$h_0(x)$ & $1$ & $0$ & $0$ & $0$ \\
\hline
$h_1(x)$ & $0$ & $1$ & $0$ & $0$ \\
\hline
$H_0(x)$ & $0$ & $0$ & $1$ & $0$ \\
\hline
$H_1(x)$ & $0$ & $0$ & $0$ & $1$ \\
\hline
\end{tabular}
\end{table}
}
\frame{
\frametitle{构造函数$h_0(x)$}
当$x = x_1$时,函数$h_0(x)$的值为$0$,
同时,函数$h_0(x)$的导数也为$0$,
\begin{center}
$\Downarrow$
\end{center}
函数$h_0(x)$必有因子$(x-x_1)^2$。
\begin{center}
$\Downarrow$
\end{center}
另外,$h_0(x)$是一个不超过三次的代数多项式,
\begin{center}
$\Downarrow$
\end{center}
\begin{equation*}
h_0(x) = [a + b(x-x_0)] \left( \frac{x - x_1}{x_0 - x_1} \right)^2
\end{equation*}
其中$a$和$b$为待定系数。
}
\frame{
\frametitle{求解$h_0(x)$}
由$h_0(x_0) = 1$可得$a=1$。
\begin{equation*}
h'_0(x) = b \left( \frac{x - x_1}{x_0 - x_1} \right)^2 + [1 + b (x-x_0)] \times 2 \times \left( \frac{x - x_1}{x_0 - x_1} \right)^ \times \frac{1}{x_0 - x_1}
\end{equation*}
$h'_0(x_0) = 0$
\begin{center}
$\Downarrow$
\end{center}
\begin{equation*}
b = \frac{2}{x_1 - x_0}
\end{equation*}
\begin{center}
$\Downarrow$
\end{center}
\begin{equation*}
h_0 (x) = \left ( 1 + 2 \frac{x- x_0}{x_1 - x_0} \right) \left ( \frac{x-x_1}{x_0 - x_1} \right)^2
\end{equation*}
同理可得
\begin{equation*}
h_1 (x) = \left ( 1 + 2 \frac{x- x_1}{x_0 - x_1} \right) \left ( \frac{x-x_0}{x_1 - x_0} \right)^2
\end{equation*}
}
\frame{
\frametitle{构造函数$H_0(x)$}
由于$H_0(x)$在$x_0$处的函数值为$0$,在$x_1$处不仅函数值为$0$,而且微商也为$0$
\begin{center}
$\Downarrow$
\end{center}
$H_0(x)$必有因子$(x - x_0)(x - x_1)^2$
\begin{center}
$\Downarrow$
\end{center}
又因为$H_0(x)$是一个不超过三次的代数多项式,而$(x - x_0)(x - x_1)^2$已经是一个三次多项式,从而$H_0(x)$可表示为:
\begin{equation*}
H_0(x) = c (x - x_0) \left( \frac{x - x_1}{x_0-x_1} \right)^2
\end{equation*}
其中$c$为待定系数。
}
\frame{
\frametitle{求解$H_0(x_0) $}
对$H_0(x_0)$求微商,可得:
\begin{equation*}
H'_0(x) = c \left[ \left( \frac{x - x_1}{x_0-x_1} \right)^2 + 2 (x - x_0) \frac{X-X_1}{(x_0 - x_1)^2} \right]
\end{equation*}
\begin{center}
$\Downarrow$
\end{center}
由$H'_0(x) = 1$,可得$c=1$,于是:
\begin{equation*}
H_0(x) = (x - x_0) \left( \frac{x - x_1}{x_0-x_1} \right)^2
\end{equation*}
\begin{center}
$\Downarrow$
\end{center}
同理可得:
\begin{equation*}
H_1(x) = (x - x_1) \left( \frac{x - x_0}{x_1-x_0} \right)^2
\end{equation*}
}
\frame{
综上所述,我们可得到三次Hermite插值公式
\begin{equation*}
\begin{array}{l c l}
H(x) & = & y_0 h_0(x) + y_1 h_1(x) + m_0 H_0(x) + m_1 H_1(x) \\
& = & \left ( 1 + 2 \frac{x- x_0}{x_1 - x_0} \right) \left ( \frac{x-x_1}{x_0 - x_1} \right)^2 y_0 + \left ( 1 + 2 \frac{x- x_1}{x_0 - x_1} \right) \left ( \frac{x-x_0}{x_1 - x_0} \right)^2 y_1\\
& & + (x - x_0) \left( \frac{x - x_1}{x_0-x_1} \right)^2 m_0 + (x - x_1) \left( \frac{x - x_0}{x_1-x_0} \right)^2 m_1 \\
\end{array}
\end{equation*}
}
\frame{
\begin{block}{定理}
设$H(x)$是$f(x)$的三次Hermite插值函数,$[a, b]$是包含$x_0$,$x_1$的任一区间,
并设$f(x) \in C^3 [a, b]$,$f^{(4)}(x)$在 $[a, b]$上存在,则对任意给定的$x \in [a, b]$,总存在一点$\xi \in [a, b]$,使得
\begin{equation*}
R(x) = f(x) - H(x) = \frac{f^{(4)}(\xi)}{4!} (x - x_0)^2 (x -x_1)^2
\end{equation*}
其中$\xi$依赖于$x$。
\end{block}
}
\frame{
\frametitle{$2n+1$次Hermite插值}
给定$y = f(x)$的函数表,
\begin{table}
\begin{tabular}{| c | c | c | c | c |}
\hline
$x$ & $x_0$ & $x_1$ & $\cdots$ & $x_n$ \\
\hline
$y = f(x)$ & $y_0$ & $y_1$ & $\cdots$ & $y_n$ \\
\hline
$y' = f'(x)$ & $m_0$ & $m_1$ & $\cdots$ & $m_n$ \\
\hline
\end{tabular}
\end{table}
构造函数$H(x)$满足条件:
\begin{itemize}
\item $H(x)$为不超过$2n+1$次的代数多项式
\item $H(x_i) = y_i$,$H'(x_i) = m_i$ ($i = 0, 1, \ldots, n$)
\end{itemize}
则称$H(x)$为$2n+1$次Hermite插值函数。
}
\frame{
设
\begin{equation*}
H(x) = \sum_{i=0}^n \left[ y_i h_i (x) + m_i H_i(x) \right]
\end{equation*}
$i = 0, 1, \ldots, n$,其中$h_i(x)$和$H_i(x)$为不超过$2n+1$的代数多项式,且满足
\begin{table}
\begin{tabular}{| c | c c c c| c c c c|}
\hline
& & 函数值 & & & & 导数值 & & \\
& $x_0$ & $x_1$ & $\cdots$ & $x_n$ & $x_0$ & $x_1$ & $\cdots$ & $x_n$ \\
\hline
$h_0(x)$ & $1$ & $0$ & $\cdots$ & $0$ & $0$ & $0$ & $\cdots$ & $0$ \\
\hline
$h_1(x)$ & $0$ & $1$ & $\cdots$ & $0$ & $0$ & $0$ & $\cdots$ & $0$ \\
\hline
$\vdots$ & $\vdots$ & $\vdots$ & $\vdots$ & $\vdots$ & $\vdots$ & $\vdots$ & $\vdots$ & $\vdots$ \\
\hline
$h_n(x)$ & $0$ & $0$ & $\cdots$ & $1$ & $0$ & $0$ & $\cdots$ & $0$ \\
\hline
$H_0(x)$ & $0$ & $0$ & $\cdots$ & $0$ & $1$ & $0$ & $\cdots$ & $0$ \\
\hline
$H_1(x)$ & $0$ & $0$ & $\cdots$ & $0$ & $0$ & $1$ & $\cdots$ & $0$ \\
\hline
$\vdots$ & $\vdots$ & $\vdots$ & $\vdots$ & $\vdots$ & $\vdots$ & $\vdots$ & $\vdots$ & $\vdots$ \\
\hline
$H_n(x)$ & $0$ & $0$ & $\cdots$ & $0$ & $0$ & $0$ & $\cdots$ & $1$ \\
\hline
\end{tabular}
\end{table}
}
\frame{
\frametitle{构造$h_0(x)$}
$h_0(x) = h'_0(x) = 0$
\begin{center}
$\Downarrow$
\end{center}
$h_0(x)$必有因子$[(x-x_1)(x-x_2)\cdots(x-x_n)]^2$。
\begin{center}
$\Downarrow$
\end{center}
$h_0(x)$为不超过$2n+1$次的代数多项式,
\begin{center}
$\Downarrow$
\end{center}
\begin{equation*}
h_0(x) = (ax + b) L_0^2(x)
\end{equation*}
其中$a$和$b$为待定常数;
\begin{equation*}
L_0(x) = \frac{(x-x_1)(x-x_2)\cdots(x-x_n)}{(x_0-x_1)(x_0-x_2)\cdots(x_0-x_n)}
\end{equation*}
}
\frame{
因为$h_0(x_0) =1$,故有$a x_0 + b = 1$,即$b = 1 -a x_0$。
\begin{center}
$\Downarrow$
\end{center}
对$h_0(x)$求一介导数,可得:
\begin{equation*}
h'_0(x) = a L_0^2(x) + 2 (ax+b) L_0(x) L'_0(x)
\end{equation*}
由$h'_0(x_0) = 0$,故有:
\begin{equation*}
h'_0(x_0) = a + 2 L'_0(x) = 0
\end{equation*}
\begin{center}
$\Downarrow$
\end{center}
\begin{equation*}
a = -2 L'_0 (x_0)
\end{equation*}
\begin{equation*}
b = 1 + 2 x_0 L'_0 (x_0)
\end{equation*}
\begin{equation*}
h_0(x) = [1 - 2 (x - x_0) L'_0 (x_0)] L^2_0(x)
\end{equation*}
}
\frame{
为求$L'_0(x)$,对$L_0(x)$两边取对数可得
\begin{equation*}
\begin{array}{l l}
\ln [L_0(x)] = & \ln(x -x_1) + \ln(x -x_2) + \cdots + \ln(x -x_n) \\
& - \left[ \ln(x_0 -x_1) + \ln(x_0 -x_2) + \dots + \ln(x_0 -x_n) \right] \\
\end{array}
\end{equation*}
对上式两边求导数可得:
\begin{equation*}
\frac{1}{L_0(x)} L'_0(x) = \sum_{j=1}^n \frac{1}{x - x_j}
\end{equation*}
\begin{center}
$\Downarrow$
\end{center}
\begin{equation*}
L'_0(x) = L_0(x) \left( \sum_{j=1}^n \frac{1}{x - x_j} \right)
\end{equation*}
由此可得
\begin{equation*}
L'_0(x_0) = \sum_{j=1}^n \frac{1}{x_0 - x_j}
\end{equation*}
}
\frame{
可以得到:
\begin{equation*}
h_0(x) = \left[ 1 - 2 (x -x_0) \left( \sum_{j=1}^n \frac{1}{x_0 - x_j} \right) \right] L_0^2(x)
\end{equation*}
\vspace{5mm}
用同样的方法可以得到:
\begin{equation*}
h_i(x) = \left[ 1 - 2 (x -x_i) \left( \sum_{j=1, j\neq i}^n \frac{1}{x_i - x_j} \right) \right] L_i^2(x)
\end{equation*}
其中$i = 0, 1, \ldots, n$
}
\frame{
\frametitle{构造$H_0(x)$}
$H_0(x_0) = 0$和$H_0(x_i) = H'_0(x_i) =0$,其中$i = 0, 1, \ldots, n$
\begin{center}
$\Downarrow$
\end{center}
$H_0(x)$必有因子$(x - x_0) [(x-x_1)(x-x_2)\cdots(x-x_n)]^2$。
\begin{center}
$\Downarrow$
\end{center}
$H_0(x)$为不超过$2n+1$次的代数多项式,
\begin{center}
$\Downarrow$
\end{center}
\begin{equation*}
H_0(x) = a (x - x_0) L_0^2(x)
\end{equation*}
其中$a$为待定常数;
}
\frame{
对$H_0(x)$求一介导数,可得:
\begin{equation*}
H'_0(x) = a \left[ L_0^2(x) + 2 (x-x_0) L_0(x) L'_0(x) \right]
\end{equation*}
由$H'_0(x) = 1$可得$a=1$。
于是可以得到:
\begin{equation*}
H_0(x) = (x - x_0) L_0^2(x)
\end{equation*}
用同样的方法可以得到:
\begin{equation*}
H_i(x) = (x - x_i) L_i^2(x)
\end{equation*}
其中$i = 0, 1, \ldots, n$
}
\frame{
$2n+1$次Hermite插值为
\begin{equation*}
\begin{array}{r l}
H(x) = & \sum_{i=0}^n \left[ y_i h_i (x) + m_i H_i(x) \right] \\
= & \sum_{i=0}^n \left\{ y_i \left[ 1 - 2 (x -x_i) \left( \sum_{j=1, j\neq i}^n \frac{1}{x_i - x_j} \right) \right] L_i^2(x) + m_i (x - x_i) L_i^2(x) \right\} \\
\end{array}
\end{equation*}
\begin{block}{定理}
设$H(x)$是$f(x)$的$2n+1$次Hermite插值函数,$[a, b]$是包含$x_0$,$x_1$,$\cdots$,$x_n$的任一区间,
并设$f(x) \in C^{2n+1} [a, b]$,$f^{(2n+2)}(x)$在 $[a, b]$上存在,则对任意给定的$x \in [a, b]$,总存在一点$\xi \in [a, b]$,使得
\begin{equation*}
R(x) = f(x) - H(x) = \frac{f^{(2n+2)}(\xi)}{(2n+2)!} \omega^2(x)
\end{equation*}
其中$\xi$依赖于$x$,$\omega(x) = (x-x_0)(x-x_1)\cdots(x-x_n)$。
\end{block}
}
\subsection{分段三次Hermite插值}
\frame{
给定函数$y = f(x)$ 的函数表
\begin{table}
\begin{tabular}{| c | c | c | c |}
\hline
$x$ & $x_0$ & $\cdots$ & $x_n$ \\
\hline
$y = f(x)$ & $y_0$ & $\cdots$ & $y_n$ \\
\hline
\end{tabular}
\end{table}
构造函数$q(x)$满足条件:
\begin{itemize}
\item $q(x)$在$[x_i, x_{i+1}]$($i = 0, 1, \ldots, n-1$)上为三次代数多项式
\item $q(x_i) = y_i$,$q'(x_i) = m_i$($i = 0, 1, \ldots, n$)
\item $q(x) \in C^1[a, b]$,其中$a = f(x_0)$,$b = f(x_n)$
\end{itemize}
则$q(x)$称为分段三次Hermite插值函数。
其在区间$[x_i, x_{i+1}]$上有
\begin{equation*}
\begin{array}{l c l}
q(x) & = & \left ( 1 + 2 \frac{x- x_i}{x_{i+1} - x_i} \right) \left ( \frac{x-x_{i+1}}{x_i - x_{i+1}} \right)^2 y_i + \left ( 1 + 2 \frac{x- x_{i+1}}{x_i - x_{i+1}} \right) \left ( \frac{x-x_i}{x_{i+1} - x_i} \right)^2 y_{i+1} \\
& & + (x - x_i) \left( \frac{x - x_{i+1}}{x_i-x_{i+1}} \right)^2 m_i + (x - x_{i+1}) \left( \frac{x - x_i}{x_{i+1}-x_i} \right)^2 m_{i+1} \\
\end{array}
\end{equation*}
}
\frame{
分段三次Hermite插值函数$q(x)$可由插值基函数组合而成:
若在插值区间$[a,b]$上定义一组分段三次Hermite插值基函数$h_i(x)$和$H_i(x)$($i = 0, 1, \ldots, n$),则可得
\begin{equation*}
q(x) = \sum_{i=0}^n [y_i h_i(x) + m_i H_i(x)]
\end{equation*}
其中$h_i(x)$和$H_i(x)$分别为
\begin{equation*}
h_0(x) = \left\{
\begin{array}{c c}
\left( 1 + 2 \frac{x-x_0}{x_1-x_0} \right) \left( \frac{x-x_1}{x_0-x_1} \right)^2 & (x_0 \le x \le x_1) \\
0 & (x_1 < x \le x_n)\\
\end{array}
\right.
\end{equation*}
\begin{equation*}
h_i(x) = \left\{
\begin{array}{c c}
\left( 1 + 2 \frac{x-x_i}{x_{i-1}-x_i} \right) \left( \frac{x-x_{i-1}}{x_i-x_{i-1}} \right)^2 & (x_{i-1} \le x \le x_i) \\
\left( 1 + 2 \frac{x-x_i}{x_{i+1}-x_i} \right) \left( \frac{x-x_{i+1}}{x_i-x_{i+1}} \right)^2 & (x_i < x \le x_{i+1}) \\
0 & [a, b] - [x_{i-1}, x_{i+1}]\\
\end{array}
\right.
\end{equation*}
\begin{equation*}
h_n(x) = \left\{
\begin{array}{c c}
\left( 1 + 2 \frac{x-x_n}{x_{n-1}-x_n} \right) \left( \frac{x-x_{n-1}}{x_n-x_{n-1}} \right)^2 & (x_{n-1} \le x \le x_n) \\
0 & (x_0 \le x < x_{n-1})\\
\end{array}
\right.
\end{equation*}
}
\frame{
\begin{equation*}
H_0(x) = \left\{
\begin{array}{c c}
(x -x_0) \left( \frac{x-x_1}{x_0-x_1} \right)^2 & (x_0 \le x \le x_1) \\
0 & (x_1 < x \le x_n)\\
\end{array}
\right.
\end{equation*}
\begin{equation*}
H_i(x) = \left\{
\begin{array}{c c}
(x -x_i) \left( \frac{x-x_{i-1}}{x_i-x_{i-1}} \right)^2 & (x_{i-1} \le x \le x_i) \\
(x -x_i) \left( \frac{x-x_{i+1}}{x_i-x_{i+1}} \right)^2 & (x_i < x \le x_{i+1}) \\
0 & [a, b] - [x_{i-1}, x_{i+1}]\\
\end{array}
\right.
\end{equation*}
\begin{equation*}
H_n(x) = \left\{
\begin{array}{c c}
(x -x_n) \left( \frac{x-x_{n-1}}{x_n-x_{n-1}} \right)^2 & (x_{n-1} \le x \le x_n) \\
0 & (x_0 \le x < x_{n-1})\\
\end{array}
\right.
\end{equation*}
}
\frame{
\begin{block}{定理}
设如下表所示的函数表,令$a=x_0$,$b=x_n$,$f(x) \in C^{(3)} [a, b]$,$f^{(4)}(x)$在 $[a, b]$上存在,$q(x)$是$f(x)$的分段三次Hermite插值函数,则有
\begin{equation*}
| R(x) | = | f(x) - q(x) | \le \frac{h^4}{384} M
\end{equation*}
其中
\begin{equation*}
h= \max_{0 \le i \le n-1} |x_{i+1} - x_i|
\end{equation*}
\begin{equation*}
M = \max_{0 \le i \le n-1} |f^{(4)}(x)|
\end{equation*}
\end{block}
\begin{table}
\begin{tabular}{| c | c | c | c | c |}
\hline
$x$ & $x_0$ & $x_1$ & $\cdots$ & $x_n$ \\
\hline
$y = f(x)$ & $y_0$ & $y_1$ & $\cdots$ & $y_n$ \\
\hline
$y' = f'(x)$ & $m_0$ & $m_1$ & $\cdots$ & $m_n$ \\
\hline
\end{tabular}
\end{table}
}